# Applied Nonlinear Control

Nguyen Tan Tien - 2002.3

_________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

C.1 Introduction
Why Nonlinear Control ? Nonlinear control is a mature subject with a variety of powerful methods and a long history of successful industrial applications ⇒ Why so many researchers have recently showed an active interest in the development and applications of nonlinear control methodologies ?

Improvement of existing control systems Linear control methods rely on the key assumption of small range operation for the linear model to be valid. When the required operation range is large, a linear controller is likely to perform very poorly or to be unstable, because the nonlinearities in the system cannot be properly compensated for. Nonlinear controllers may handle the nonlinearities in large range operation directly. Ex: pendulum

Analysis of hard nonlinearities One of the assumptions of linear control is that the system model is indeed linearizable. However, in control systems, there are many nonlinearities whose discontinuous nature does not allow linear approximation. Ex: Coulomb friction, backlash

Dealing with model uncertainties In designing linear controllers, it is usually necessary to assume that the parameters of the system model are reasonably well known. However in many control problems involve uncertainties in the model parameters. Nonlinearities can be intentionally introduced into the control part of a control system so that model uncertainties can be tolerated. Two classes of nonlinear controllers for this purpose are robust controllers and adaptive controllers. Ex: parameter variations

Design simplicity Good nonlinear control designs may be simpler and more intuitive than their linear counterparts. Ex: & x = Ax + Bu & x = f + gu

___________________________________________________________________________________________________________ 1 Chapter 1 Introduction

Applied Nonlinear Control

Nguyen Tan Tien - 2002.3

_________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

2. Phase Plane Analysis
Phase plane analysis is a graphical method for studying second-order systems. This chapter’s objective is to gain familiarity of the nonlinear systems through the simple graphical method. 2.1 Concepts of Phase Plane Analysis 2.1.1 Phase portraits The phase plane method is concerned with the graphical study of second-order autonomous systems described by & x1 = f1 ( x1 , x 2 ) & 2 = f 2 ( x1 , x 2 ) x where x1 , x 2 : states of the system f1 , f 2 : nonlinear functions of the states Geometrically, the state space of this system is a plane having x1 , x 2 as coordinates. This plane is called phase plane. The solution of (2.1) with time varies from zero to infinity can be represented as a curve in the phase plane. Such a curve is called a phase plane trajectory. A family of phase plane trajectories is called a phase portrait of a system. Example 2.1 Phase portrait of a mass-spring system_______
& x

The nature of the system response corresponding to various initial conditions is directly displayed on the phase plane. In the above example, we can easily see that the system trajectories neither converge to the origin nor diverge to infinity. They simply circle around the origin, indicating the marginal nature of the system’s stability. A major class of second-order systems can be described by the differential equations of the form && = f ( x, x) & x (2.3)

(2.1a) (2.1b)

In state space form, this dynamics can be represented & with x1 = x and x 2 = x as follows & x1 = x2 & x2 = f ( x1 , x 2 ) 2.1.2 Singular points A singular point is an equilibrium point in the phase plane. Since an equilibrium point is defined as a point where the & system states can stay forever, this implies that x = 0 , and using (2.1)  f1 ( x1 , x 2 ) = 0   f 2 ( x1 , x 2 ) = 0 (2.4)

For a linear system, there is usually only one singular point although in some cases there can be a set of singular points.
x

k =1
0

Example 2.2 A nonlinear second-order system____________
& x
9

m =1
(a ) (b)
-6 -3

6
convergence 3 area

Fig. 2.1 A mass-spring system and its phase portrait The governing equation of the mass-spring system in Fig. 2.1 is the familiar linear second-order differential equation

3

6

x

unstable

&& + x = 0 x

(2.2)

-3

Assume that the mass is initially at rest, at length x0 . Then the solution of this equation is x(t ) = x0 cos(t ) & x(t ) = − x0 sin(t ) Eliminating time t from the above equations, we obtain the equation of the trajectories
2 & x + x 2 = x0 2

-6
to infinity

divergence area

-9

Fig. 2.2 A mass-spring system and its phase portrait

& Consider the system && + 0.6 x + 3x + x 2 = 0 whose phase x portrait is plot in Fig. 2.2.
The system has two singular points, one at (0,0) and the other at (−3,0) . The motion patterns of the system trajectories in the vicinity of the two singular points have different natures. The trajectories move towards the point x = 0 while moving away from the point x = −3 .
__________________________________________________________________________________________

This represents a circle in the phase plane. Its plot is given in Fig. 2.1.b.
__________________________________________________________________________________________

___________________________________________________________________________________________________________ 1 Chapter 2 Phase Plane Analysis

Applied Nonlinear Control

Nguyen Tan Tien - 2002.3

_________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Why an equilibrium point of a second order system is called a singular point ? Let us examine the slope of the phase portrait. The slope of the phase trajectory passing through a point ( x1 , x 2 ) is determined by dx2 f (x , x ) = 2 1 2 dx1 f1 ( x1 , x 2 ) (2.5)

f ( x1 , x2 ) = f ( x1 ,− x 2 ) f ( x1 , x2 ) = − f ( x1 ,− x2 )

⇒ symmetry about the x1 axis.

⇒ symmetry about the x2 axis.

f ( x1 , x2 ) = − f (− x1 ,− x 2 ) ⇒ symmetry about the origin. 2.2 Constructing Phase Portraits There are a number of methods for constructing phase plane trajectories for linear or nonlinear system, such that so-called analytical method, the method of isoclines, the delta method, Lienard’s method, and Pell’s method. Analytical method There are two techniques for generating phase plane portraits analytically. Both technique lead to a functional relation between the two phase variables x1 and x2 in the form g ( x1 , x 2 ) = 0 (2.6)

where f1 , f 2 are assumed to be single valued functions. This implies that the phase trajectories will not intersect. At singular point, however, the value of the slope is 0/0, i.e., the slope is indeterminate. Many trajectories may intersect at such point, as seen from Fig. 2.2. This indeterminacy of the slope accounts for the adjective “singular”. Singular points are very important features in the phase plane. Examining the singular points can reveal a great deal of information about the properties of a system. In fact, the stability of linear systems is uniquely characterized by the nature of their singular points. Although the phase plane method is developed primarily for second-order systems, it can also be applied to the analysis of first-order systems of the form & x + f ( x) = 0 The difference now is that the phase portrait is composed of a single trajectory. Example 2.3 A first-order system_______________________

where the constant c represents the effects of initial conditions (and, possibly, of external input signals). Plotting this relation in the phase plane for different initial conditions yields a phase portrait. The first technique involves solving (2.1) for x1 and x2 as a function of time t , i.e., x1 (t ) = g1 (t ) and x2 (t ) = g 2 (t ) , and then, eliminating time t from these equations. This technique was already illustrated in example 2.1. The second technique, on the other hand, involves directly dx f (x , x ) eliminating the time variable, by noting that 2 = 2 1 2 dx1 f1 ( x1 , x 2 ) and then solving this equation for a functional relation between x1 and x2 . Let us use this technique to solve the massspring equation again. Example 2.4 Mass-spring system_______________________

& Consider the system x = −4 x + x there are three singular
points, defined by − 4 x + x 3 = 0 , namely, x = 0, − 2, 2 . The phase portrait of the system consists of a single trajectory, and is shown in Fig. 2.3.
& x

3

unstable
-2

stable

unstable
x

0

2

& By noting that && = (dx / dx) /( dx / dt ) , we can rewrite (2.2) as x & dx 2 & & x + x = 0 . Integration of this equation yields x 2 + x 2 = x0 . dx
__________________________________________________________________________________________

Fig. 2.3 Phase trajectory of a first-order system The arrows in the figure denote the direction of motion, and whether they point toward the left or the right at a particular & point is determined by the sign of x at that point. It is seen from the phase portrait of this system that the equilibrium point x = 0 is stable, while the other two are unstable.
__________________________________________________________________________________________

Most nonlinear systems cannot be easily solved by either of the above two techniques. However, for piece-wise linear systems, an important class of nonlinear systems, this can be conveniently used, as the following example shows. Example 2.5 A satellite control system___________________
Jets
θd = 0
U -U

u

1 p

Sattellite θ&

1 p

θ

2.1.3 Symmetry in phase plane portrait

& x Let us consider the second-order dynamics (2.3): && = f ( x, x) . The slope of trajectories in the phase plane is of the form
dx2 f ( x1 , x2 ) =− & dx1 x

Fig. 2.4 Satellite control system

Fig. 2.4 shows the control system for a simple satellite model. The satellite, depicted in Fig. 2.5.a, is simply a rotational unit inertia controlled by a pair of thrusters, which can provide either a positive constant torque U (positive firing) or negative torque (negative firing). The purpose of the control system is Since symmetry of the phase portraits also implies symmetry to maintain the satellite antenna at a zero angle by of the slopes (equal in absolute value but opposite in sign), we appropriately firing the thrusters. can identify the following situations: ___________________________________________________________________________________________________________ 2 Chapter 2 Phase Plane Analysis

Applied Nonlinear Control

Nguyen Tan Tien - 2002.3

_________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

& The mathematical model of the satellite is θ& = u , where u is the torque provided by the thrusters and θ is the satellite angle. Let us examine on the phase plane the behavior of the control system when the thrusters are fired according to the control law −U u (t ) =  U if if

The method of isoclines (ñöôø ng ñaú ng khuynh) The basic idea in this method is that of isoclines. Consider the & & dynamics in (2.1): x1 = f1 ( x1 , x 2 ) and x2 = f 2 ( x1 , x 2 ) . At a point ( x1 , x2 ) in the phase plane, the slope of the tangent to the trajectory can be determined by (2.5). An isocline is defined to be the locus of the points with a given tangent slope. An isocline with slope α is thus defined to be dx2 f (x , x ) = 2 1 2 =α dx1 f1 ( x1 , x 2 ) This is to say that points on the curve f 2 ( x1 , x 2 ) = α f1 ( x1 , x2 ) all have the same tangent slope α . In the method of isoclines, the phase portrait of a system is generated in two steps. In the first step, a field of directions of tangents to the trajectories is obtained. In the second step, phase plane trajectories are formed from the field of directions. Let us explain the isocline method on the mass-spring system in (2.2): && + x = 0 . The slope of the trajectories is easily seen x to be dx2 x =− 1 dx1 x2

θ >0 θ <0

(2.7)

which means that the thrusters push in the counterclockwise direction if θ is positive, and vice versa. As the first step of the phase portrait generation, let us consider the phase portrait when the thrusters provide a & positive torque U . The dynamics of the system is θ& = U , & dθ& = U dθ . Therefore, the phase portrait which implies that θ trajectories are a family of parabolas defined by θ& 2 = 2U θ + c1 , where c1 is constant. The corresponding phase portrait of the system is shown in Fig. 2.5.b. When the thrusters provide a negative torque −U , the phase trajectories are similarly found to be θ& 2 = −2U x + c1 , with the corresponding phase portrait as shown in Fig. 2.5.c.

θ

antenna

& x
x

& x
x

Therefore, the isocline equation for a slope α is x1 + α x 2 = 0

u

u =U

u = −U

Fig. 2.5 Satellite control using on-off thrusters The complete phase portrait of the closed-loop control system can be obtained simply by connecting the trajectories on the left half of the phase plane in 2.5.b with those on the right half of the phase plane in 2.5.c, as shown in Fig. 2.6. parabolic trajectories
& x

i.e., a straight line. Along the line, we can draw a lot of short line segments with slope α . By taking α to be different values, a set of isoclines can be drawn, and a field of directions of tangents to trajectories are generated, as shown in Fig. 2.7. To obtain trajectories from the field of directions, we assume that the tangent slopes are locally constant. Therefore, a trajectory starting from any point in the plane can be found by connecting a sequence of line segments.

α =1

& x

α = −1

x

x

α =∞
u = +U u = −U switching line

Fig.2.6 Complete phase portrait of the control system The vertical axis represents a switching line, because the control input and thus the phase trajectories are switched on that line. It is interesting to see that, starting from a nonzero initial angle, the satellite will oscillate in periodic motions under the action of the jets. One can concludes from this phase portrait that the system is marginally stable, similarly to the mass-spring system in Example 2.1. Convergence of the system to the zero angle can be obtained by adding rate feedback.

Fig. 2.7 Isoclines for the mass-spring system Example 2.6 The Van der Pol equation__________________ For the Van der Pol equation && + 0.2( x 2 − 1) x + x = 0 & x an isocline of slope α is defined by

& & dx 0.2( x 2 − 1) x + x =− =α & dx x __________________________________________________________________________________________ ___________________________________________________________________________________________________________ 3 Chapter 2 Phase Plane Analysis

the length of time corresponding to the & increment ∆x is ∆t ≈ ∆x / x . This implies that. 2. and then plots x with respects to t and x with respects to t . we solve for the time history x(t ) = k1e λ1 t + k 2 e λ2 t x(t ) = k1e λ1 t for λ1 ≠ λ2 for λ1 = λ2 (2. ∫ λ1 . in order to obtain the time corresponding to the motion from one point to another point along the trajectory.10) isoclines Fig.2002. From (2. λ2 are complex conjugates with real parts equal to 0 We now briefly discuss each of the above four cases Stable or unstable node (Fig. Because of the unstable pole λ2 .9a) and then substitution of & (2. as plot in Fig. differentiation of (2. This closed curve corresponds to a limit cycle. Saddle point (Fig.2( x 2 − 1) x + x + α x = 0 all have the same slope α .Applied Nonlinear Control Nguyen Tan Tien . the points on the curve & & 0.9b) α =1 trajectory -2 2 α = −5 x1 limit cycle Transform these equations into a scalar second-order & & differential equation in the form b x 2 = b c x1 + d ( x1 − a x1 ) .8) whre the constant λ1 . Therefore. Therefore.c) The second case ( λ1 < 0 < λ2 ) corresponds to a saddle point. However. 2. We now to describe two techniques for computing time history from phase portrait. λ2 are complex conjugates with non-zero real parts • λ1 . depending on the values of a and b . we x will simply consider the second-order linear system described by && + a x + b x = 0 & x (2. There is no oscillation in the trajectories. α = 0 α = −1 x2 where x corresponding to time t and x0 corresponding to time t 0 . This implies that. 2.11a) (2.8).9. almost all of the system trajectories diverge to infinity. λ2 > 0 : singularity point is called unstable node. we should divide the corresponding part of the trajectory into a number of small segments (not necessarily equally spaced). the change of x is approximately & ∆x ≈ x ∆t (2. By taking α of different isoclines can be obtained. find the time associated with each segment. as shown in the plot. 2. λ2 < 0 : singularity point is called stable node. the trajectories in the vicinity of this singularity point can display quite different characteristics. then the area under the resulting curve is the corresponding time interval. Both of techniques involve a step-by step procedure for recovering time. as will be discussed further in section 2.a -b) The first case corresponds to a node. & t − t 0 ≈ (1 / x) dx x0 For linear systems described by (2. namely the origin. and then add up the results.9a) (2.9b) leads to &&1 = (a + d ) x1 + (c b − a d ) x1 . we simply compute the time t for each point on the & phase trajectory.9.8 Phase portrait of the Van der Pol equation Short line segments are drawn on the isoclines to generate a field of tangent directions.4 Phase Plane Analysis of Linear Systems The general form of a linear second-order system is & x1 = a x1 + b x2 & x2 = c x1 + d x 2 (2. we can write dt = dx / x . Consequently. λ2 are both real and have the same sign (+ or -) • λ1 .8. The following cases can occur • λ1 . & Obtaining time from t ≈ (1 / x) dx & & Since x = dx / dt . if we plot a phase plane portrait & with new coordinates x and (1 / x ) .5. To obtain the history of states corresponding to a certain initial condition.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Therfore.10).11b) + k2 t e λ2 t 2.3 Determining Time from Phase Portraits Time t does not explicitly appear in the phase plane having x1 and x 2 as coordinates. λ2 are the solutions of the characteristic equation s 2 + as + b = ( s − λ1 )( s − λ2 ) = 0 The roots λ1 . there is only one singular point (b ≠ 0) . & Obtaining time from ∆t ≈ ∆x / x In a short time ∆t . λ2 are both real and have opposite sign • λ1 . A node can be stable or unstable: λ1 . __________________________________________________________________________________________ To obtain the phase portrait of this linear system. ∫ x ___________________________________________________________________________________________________________ 4 Chapter 2 Phase Plane Analysis . 2. It is interesting to note that there exists a closed curved in the portrait. The phase portraits can be obtained. λ2 can be explicitly represented as λ1 = − a + a 2 − 4b − a − a 2 − 4b and λ2 = 2 2 & where x is the velocity corresponding to the increment ∆x . and the trajectories starting from both outside and inside converge to this curve.

2. Using Taylor expansion.12) In discussing the phase plane analysis of nonlinear system. 2.2002. 2. This. we may simply consider Eq.10. and therefore. is not true for nonlinear systems. 2 2  x1 = x 2 − x1 ( x1 + x 2 − 1) 2 • Nonlinear systems can display much more complicated & (c)  (2. g 2 contain higher order terms.13) 2 2 nonlinear systems can be approximated by the behavior  x2 = − x1 + x 2 ( x1 + x2 − 1) & of a linear system. 2.10. the local behavior of the nonlinear system can be approximated by the patterns shown in Fig.a). two points should be kept in mind: • Phase plane analysis of nonlinear systems is related to 2 2  x = x 2 + x1 ( x1 + x 2 − 1) & that of liner systems. The trajectory has to be both closed. we can shift the singular point to the origin. 2. x 2 ) where g1 . x jω stable focus & x σ (d ) jω unstable focus & x x σ (e) Limit cycle In the phase plane. λ2 ) > 0 : unstable focus Center point (Fig. Re(λ1 .10. we can distinguish three kinds of limit cycles. such as multiple equilibrium 2 2  x 2 = − x1 − x2 ( x1 + x2 − 1) 2 & points and limit cycles.Applied Nonlinear Control Nguyen Tan Tien .9. unstable. without loss of generality. 2. • Stable Limit Cycles: all trajectories in the vicinity of the limit cycle converge to it as t → ∞ (Fig. • Unstable Limit Cycles: all trajectories in the vicinity of the limit cycle diverge to it as t → ∞ (Fig. ___________________________________________________________________________________________________________ 5 Chapter 2 Phase Plane Analysis .f) The last case corresponds to a certain point. Depending on the motion patterns of the trajectories in the vicinity of the limit cycle. λ2 ) < 0 : stable focus Re(λ1 .9. Therefore.14) patterns in the phase plane. unstable node & x x σ (b) jω saddle point & x x σ (c ) In the vicinity of the origin. diverging x1 x1 x1 limit cycle limit cycle limit cycle (c) ⊗ Note that the stability characteristics of linear systems are uniquely determined by the nature of their singularity points. 2.9. and semi-stable limit cycles Example 2. All trajectories are ellipses and the singularity point is the centre of these ellipses.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ jω stable node & x x σ (a) jω Local behavior of nonlinear systems If the singular point of interest is not at the origin.10 Stable. because the local behavior of (b)  1 (2. indicating the limiting nature of the cycle (with near by trajectories converging or diverging from it).b) • Semi-Stable Limit Cycles: some of the trajectories in the vicinity of the limit cycle converge to it as t → ∞ (Fig. a limit cycle is defied as an isolated closed curve.d-e) The third case corresponds to a focus.7 Stable. 2. indicating the periodic nature of the motion. (2. the higher order terms can be neglected. by defining the difference between the original state and the singular point as a new set of state variables.1) can be rewritten in the form & x1 = a x1 + b x 2 + g1 ( x1 . Eqs.5 Phase Plane Analysis of Nonlinear Systems (a ) (b) Fig. unstable.1) with a singular point at 0.9 Phase-portraits of linear systems Stable or unstable locus (Fig. and isolated.c) x2 converging trajectories x2 diverging trajectories converging x2 jω center point & x x σ (f) Fig. the nonlinear system trajectories essentially satisfy the linearized equation & x1 = a x1 + b x2 & x2 = c x1 + d x 2 As a result. x 2 ) & x2 = c x1 + d x 2 + g 2 ( x1 . however. and semi-stable limit cycle___ Consider the following nonlinear systems 2 2  x = x2 − x1 ( x1 + x 2 − 1) & (a)  1 2 2  x 2 = − x1 − x 2 ( x1 + x 2 − 1) & (2. 2.(2.

This can also be concluded by examining the analytical solution of (2. Where. the system does not have any limit cycles any where in the phase plane.1 (Pointcare) If a limit cycle exists in the secondorder autonomous system (2. & When r > 1 . Theorem 2.Applied Nonlinear Control Nguyen Tan Tien . Therefore.2002.12) are transformed as dr = − r (r 2 − 1) dt dθ = −1 dt When the state starts on the unicycle.1). Example 2.8________________________________________ Consider the nonlinear system 2 & x1 = g ( x 2 ) + 4 x1 x 2 2 & x2 = h( x1 ) + 4 x1 x2 Since ∂ f1 ∂ f 2 2 2 + = 4( x1 + x 2 ) . no limit cycle can exist in the region Ω of the phase plane in which ∂f1 / ∂x1 + ∂f 2 / ∂x 2 does not vanish and does not change sign. Therefore. S represents the number of enclosed saddle points. then r > 0 . When r < 1 . and foci enclosed by a limit cycle.12) r (t ) = 1 1 + c0 e − 2t and θ (t ) = θ 0 − t . the above equation & shows that r (t ) = 0 . centers. __________________________________________________________________________________________ ___________________________________________________________________________________________________________ 6 Chapter 2 Phase Plane Analysis . the N=S+1.1).3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ By introducing a polar coordinates 2 2 r = x1 + x2 x  θ (t ) = tan −1  2  x   1 the dynamics of (2. then r < 0 .2 (Pointcare-Bendixson) If a trajectory of the second-order autonomous system remains in a finite region Ω . the unit circle is a stable limit cycle. N represents the number of nodes. where c0 = 1 r02 −1 Similarly. __________________________________________________________________________________________ 2. This implies that the states tend to the unit circle from outside. then one of the following is true: (a) the trajectory goes to an equilibrium point (b) the trajectory tends to an asymptotically stable limit cycle (c) the trajectory is itself a limit cycle Theorem 2. which is always strictly ∂x1 ∂x 2 positive (except at the origin). This theorem is sometime called index theorem. This implies that the state tends to the circle from inside. the state will circle around the & origin with a period 1 / 2π . we can find that the system (b) has an unstable limit cycle and system (c) has a semi-stable limit cycle.3 (Bendixson) For a nonlinear system (2.6 Existence of Limit Cycles Theorem 2.

1 The nonlinear system (3. it remains equal to x * for all future time. Physically.1 Nonlinear Systems and Equilibrium Points Nonlinear systems A nonlinear dynamic system can usually be presented by the set of nonlinear differential equations in the form & x = f ( x. these adjectives are replaced by autonomous and non-autonomous. the nominal motion trajectory. As we shall see later. for notational and analytical simplicity. The & dynamics of linear systems are of the from x = A(t ) x with A ∈ R n×n . Leting x1 = θ .1 The pendulum___________________________ (3.1 Pendulum Consider the pendulum of Fig.3).. which contained in the null-space of the matrix A. sin( x1 ) = 0. b the friction coefficient at the hinge. A linear time-invariant system & x = Ax * (3. __________________________________________________________________________________________ (3. Such a point is called an equilibrium point.3) x* (t ) x1 xn Fig. M its mass. Equilibrium points It is possible for a system trajectory to only a single point. which leads to the points (0 [2π ]. t ) where f ∈R n has a single equilibrium point (the origin 0) if A is nonsingular.2) In linear system analysis and design.1) is said to be autonomous if f does not depend explicitly on time. Let us now perturb the initial condition to be x(0) = x 0 + δ x 0 . A special class of nonlinear systems is linear system.1. we often transform the linear system equations in such a way that the equilibrium point is the origin of the state-space. 3. the subspace defined by Ax = 0. many stability problems are naturally formulated with respect to equilibrium points.e.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 3.4) . corresponding to initial condition x* (0) = x 0 .1) θ R : nonlinear vector function : state vectors : order of the system Fig. x2 x(t ) e(t ) equilibrium points) of the system if once x(t ) is equal to x * . Mathematically. i. the corresponding state-space equation is & x1 = x2 g & x2 = − x − sin x1 2 2 R MR b (3. this means that the constant vector x * satisfies 0 = f (x ) Equilibrium points can be found using (3. 0) and (π [2π ].. i. 3. Nominal motion & Let x* (t ) be the solution of x = f (x) . Definition 3. it has an infinity of equilibrium points. whose dynamics is given by the following nonlinear autonomous equation & MR 2θ& + bθ& + MgR sin θ = 0 (3. Definition 3.1) can represent both closed-loop dynamics of a feedback control system and the dynamic systems where no control signals are involved. Fundamentals of Lyapunov Theory The objective of this chapter is to present Lyapunov stability theorem and illustrate its use in the analysis and the design of nonlinear systems. 0) . A nonlinear system can have several (or infinitely many) isolated equilibrium points.. x2 = θ& . and study the associated variation of the motion error e(t ) = x(t ) − x* (t ) as illustrated in Fig. For nonlinear systems.e.6a) (3. If A is singular.2 Nominal and perturbed motions ___________________________________________________________________________________________________________ 7 Chapter 3 Fundamentals of Lyapunov Theory (3. if the system’s state equation can be written & x = f (x) Otherwise. 3. 3.e.5) x ∈ Rn n The form (3.6b) Therefore the equilibrium points are given by x2 = 0.2 A state x is an equilibrium state (or * where R is the pendulum’s length. 3. i. Example 3. the system is called non-autonomous.Applied Nonlinear Control Nguyen Tan Tien .2002.2. Autonomous and non-autonomous systems Linear systems are classified as either time-varying or timeinvariant. these points correspond to the pendulum resting exactly at the vertical up and down points. and g the gravity constant.

due to the presence of the nominal trajectory x * (t ) on the right hand side.5 An equilibrium points 0 is exponential stable if there exist two strictly positive number α and λ such that ∀t > 0. has an equilibrium point at the origin of the state space. 3. we not only want the satellite to maintain its attitude in a range determined by the magnitude of the disturbance. Assume that we slightly perturb the initial position to be x(0) = x0 + δ x0 .8) with respect to the equilibrium point 0. There is a need to estimate how fast the system trajectory approaches 0.9) means that the state vector of an exponentially stable form system converges to the origin faster than an exponential ___________________________________________________________________________________________________________ 8 Chapter 3 Fundamentals of Lyapunov Theory .3 can be written in the (3. there exist r > 0 . for any R > 0 .2002. or Lyapunov stability) means that the system trajectory can be kept arbitrarily close to the origin by starting sufficiently close to it. t ) = 0 . The resulting system trajectory is denoted as x(t ) . but also required that the attitude gradually go back to its original value. the new dynamic system. if. Lyapunov stability. we have & x * (0) = x 0 x* = f (x* ) ∀R > 0. with e as state and g in place of f. x(t ) < R or. it is still not sufficient to know that a system will converge to the equilibrium point after infinite time. states start close to 0 actually converge to 0 as time goes to infinity. stability (also called stability in the sense of Lyapunov.3 shows that the system trajectories starting form within the ball B r converge to the origin. Using the above symbols. Fig.8) with initial condition e(0) = δ x(0) . 3. instead of studying the deviation of x(t ) from x * (t ) for the original system.3 Concepts of stability Asymptotic stability and exponential stability In many engineering applications. Otherwise.unstable SR Fig.Applied Nonlinear Control Nguyen Tan Tien . The concept of exponential stability can be used for this purpose.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ & Since both x * (t ) and x(t ) are solutions of (3. t ) − f (x * .2 Concepts of Stability Notation B R : spherical region (or ball) defined by x ≤ R S R : spherical itself defined by x = R ∀ ∃ ∈ ⇒ ⇔ : for any : there exist : in the set : implies that : equivalent Stability and instability Definition 3. 33. i. ∃r > 0.2________________________________________ Consider the autonomous mass-spring system & m && + k1 x + k 2 x 3 = 0 x which contains a nonlinear term reflecting the hardening effect of the spring. Asymptotic stability means that the equilibrium is stable. This type of engineering requirement is captured by the concept of asymptotic stability. t ) (3.. However.asymptotically stable 3 1 curve 2 .3 The equilibrium state x = 0 is said to be stable if.2): x = f (x) . and in addition. this is a non-autonomous system. More formally. In many engineering applications. The geometrical implication of stability is indicated in Fig. equivalently ∀R > 0. x(t ) ∈ B r & x = f (x) x ( 0) = x 0 + δ x 0 then e(t ) satisfies the following non-autonomous differential equation & e = f (x * + e.e. (3. Example 3. we may simply study the stability of the perturbation dynamics (3. For example. Since g (0. Let us study the stability of the motion Essentially. The ball B r is called a domain of attraction of the equilibrium point. when a satellite’s attitude is disturbed from its nominal position. __________________________________________________________________________________________ 3. ∃r > 0. given that we do not want the state trajectory x(t ) to get out of a ball of arbitrarily specified radius B R .marginally stable 2 0 x(0) S r curve 3 . such that if x(0) ≤ r then x(t ) ≤ R for all t ≥ 0 . the perturbation dynamics non-autonomous system. x(t ) ≤ α x(0) e −λt in some ball B r around the origin. Therefore. t ) = g (e.4 An equilibrium points 0 is asymptotically stable if it is stable. Proceeding as before. Definition 3. Lyapunov stability is not enough. x(0) < r ⇒ ∀t ≥ 0. Definition 3. x(0) ∈ B r ⇒ ∀t ≥ 0. the definition states that the origin is stable. curve 1 . the equilibrium point is unstable. and if in addition there exist some r > 0 such that x(0) ≤ r implies that x(t ) → 0 as t → ∞ .9) x* (t ) which starts from initial point x0 . the equivalent differential equation governing the motion error e is m && + k1e + k 2 [e 3 + 3e 2 x * (t ) + 3e x *2 (t )] = 0 e Clearly. Definition 3.

Consider the autonomous system in (3. as can be seen from the system & x = − x 2 . if all eigenvalues of A are in the left-half complex plane but at least one of them is on the jω axis). as we now illustrate. we can write sin θ = sin π + cos π (θ − π ) + h.θ& = 0) is ~ && θ + ~ g ~ & θ − θ =0 R MR b 2 where f h. The origin 0 is one of the two equilibrium of this system.1 (Lyapunov’s linearization method) • If the linearized system is strictly stable (i. In practice. asymptotically stable.2). the nonlinear system is x = b x 5 .o.6 If asymptotic (or exponential) stability holds for any initial states. __________________________________________________________________________________________ 1 1 ___________________________________________________________________________________________________________ 9 Chapter 3 Fundamentals of Lyapunov Theory . its solution is x(t ) = x(0) e t − ∫0 −[1+sin 2 x (τ )] dτ . Its linearized approximation about x = 0 is & x1 = 0 + x1. and therefore x(t ) ≤ x(0) e −t . Local and global stability Definition 3..o. then the equilibrium point is asymptotically stable (for the actual nonlinear system). the system    x =0 & x = Ax (3. or unstable for the nonlinear system).t . finding a system’s linearization is often most easily done simply neglecting any term of order higher than 1 in the dynamics. x & & For example. __________________________________________________________________________________________ Example 3.12) is called the linearization (or linear approximation) of the original system at the equilibrium point 0.5________________________________________ Consider the equilibrium point (θ = π .e. It is also called globally asymptotically (or exponentially) stable.o. Assume that the control law for the original nonlinear x system has been selected to be u = sin x + x 3 + x cos 2 x . the direct method to be described can easily solve this problem.t .11) • If the linearized system is marginally stable (i. x(0) = 1 (3. Let us use the constant matrix A denote the Jacobian matrix of f with respect  ∂f  to x at x = 0: A =   ∂ x  . then one cannot conclude anything from the linear approximation (the equilibrium point may be stable. if at least one eigenvalue of A is strictly in the right-half complex plane). stands for higher-order terms in x.θ& = 0) of the pendulum in the example 3.4________________________________________ Consider the nonlinear system 2 & x1 = x2 + x1 cos x 2 & x2 = x 2 + ( x1 + 1) x1 + x1 sin x 2 Hence its linear approximation is unstable.1.e. ~ thus letting θ = θ − π . (x)    x =0 (3.2002. The linearization method fails while. the equilibrium point is said to be asymptotically (or exponentially) stable in the large. and assumed that f(x) is continuously differentiable.t. A similar procedure can be applied for a controlled system. The system can x be linearly approximated about x = 0 as && + 0 + (0 + 1) u = 0 or x && = u . whose solution is x = 1 /(1 + t ) . if all eigenvalues of A are strictly in the left-half complex plane).2).e.1 & x2 = x 2 + 0 + x1 + x1 x2 ≈ x2 + x1 1 0 & The linearized system can thus be written x =  x . Example 3.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ function. & Consider the system && + 4 x 5 + ( x 2 + 1) u = 0 .5________________________________________ & Consider the first-order system x = a x + b x 5 . Theorem 3. then the linearized closed-loop dynamics is && + x + x = 0 . as we shall see. and therefore so is the nonlinear system at this equilibrium point. It is a formalization of the intuition that a nonlinear system should behave similarly to its linearized approximation for small range motions. Indeed. But asymptotic stability does not implies guarantee exponential stability. Note that exponential stability implies asymptotic stability. then the equilibrium point is unstablle (for the nonlinear system). Since the neighborhood of θ = π .3 Linearization and Local Stability Lyapunov’s linearization method is concerned with the local stability of a nonlinear system. The application of Lyapunov’s linearization method indicate the following stability properties of the nonlinear system • a < 0 : asymptotically stable • a > 0 : unstable • a = 0 : cannot tell from the linearization & In the third case. the system x = −(1 + sin 2 x) x is exponentially convergent to x = 0 with the rate λ = 1 .t. the system’s linearization about the equilibrium point (θ = π .10) __________________________________________________________________________________________ The following result makes precise the relationship between the stability of the linear system (3. The positive number λ is called the rate of exponential convergence.Applied Nonlinear Control Nguyen Tan Tien .o. Example 3. The linearization of & this system around the origin is x = a x .. 3.2) and that of the original nonlinear system (3.. = π − θ + h. a function slower than any exponential function e − λt . • If the linearizad system is un stable (i. Then the system dynamics can be written as  ∂f  & x=  ∂ x  x + f h. Then.

3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 3. is continuously dissipated by the damper & until the mass is settled down.13) & &x & & & & & V (x) = m x && + (k 0 x + k1 x 3 ) x = x (−b x x ) = −b x 3 (3. V (x) ≤ 0 then.These ovals often called contour curves may be thought as the section of the cup by horizontal planes.e.14) and using (3. the mechanical energy.7. and if its time derivative along any state trajectory of system (3. with each oval corresponding to a positive value of Vα .4 Lyapunov’s Direct Method The basic philosophy of Lyapunov’s direct method is the mathematical extension of a fundamental physical observation: if the total energy of a mechanical (or electrical) system is continuous dissipated.2002. Positive definite functions and Lyapunov functions Definition 3. 3. is locally positive definite. projected on the ( x1 .6 A nonlinear mass-damper-spring system Total mechanical energy = kinetic energy + potential energy 1 1 k 0 x 2 + k1 x 4 2 4 0 (3.e. For instance. we may conclude the stability of a system by examining the variation of a single scalar function. indirectly reflects the magnitude of the state vector. starting from some initial value. 3. 3. ___________________________________________________________________________________________________________ 10 Chapter 3 Fundamentals of Lyapunov Theory . in a ball B R0 x≠0 ⇒ V ( x) > 0 If V (0) = 0 and the above property holds over the whole state space. (3. x 2 ) = Vα typically present a set of ovals surrounding the origin. whether linear or nonlinear.2) is & negative semi-definite. the function V (x) is positive definite and has continuous partial derivatives. Let consider the nonlinear mass-damper-spring system in Fig.7 A scalar continuous function V (x) is said to be locally positive definite if V (0) = 0 and. In 3-dimensional space. Consider a positive definite function V (x) of two state variables x1 and x2 . must eventually settle down to an equilibrium point. x 2 ) The 2-dimesional geometrical representation can be made as follows.. The rate of energy variation during the system’s motion is obtained by differentiating the first equality in (3. 3. that the stability properties of the system can be characterized by the variation of the mechanical energy of the system. then V (x) is said to be globally positive definite.8 Interpreting positive definite functions using contour curves Definition 3. i.7 Typical shape of a positive definite function V ( x1 .13) & x & x : nonlinear dissipation or damping Let us describe the geometrical meaning of locally positive definite functions.1. whose dynamic equation is 3. V (x) typically corresponds to a surface looking like an upward cup as shown in Fig. Taking x1 and x2 as Cartesian coordinates.6. The lowest point of the cup is located at the origin. The direct method of Lyapunov is based on generalization of the concepts in the above mass-spring-damper system to more complex systems. V k 0 x + k1 x 3 : nonlinear spring term nonlinear spring and damper m V = V3 V = V2 Fig. x = 0 .14) Comparing the definitions of stability and mechanical energy.15) implies that the energy of the system.8. in a ball B R0 . the function V (x) = 1 2 MR 2 x2 + MR(1 − cos x1 ) 2 which is the mechanical energy of the pendulum in Example 3. and furthermore. 3. i.1. x = 0) • assymptotic stability implies the convergence of mechanical energy to zero • instability is related to the growth of mechanical energy The relations indicate that the value of a scalar quantity.8 If.4. x 2 ) plane as shown in Fig. V (x) is said to be a Lyapunov function for the system (3.Applied Nonlinear Control Nguyen Tan Tien . & & m && + b x x + k 0 x + k1 x 3 = 0 x with b (3.15) Fig. then the system.2). V = V2 x2 V = V1 x1 0 V = V3 V3 > V2 > V1 • zero energy corresponds to the equilibrium point & (x = 0. 3. Thus.. we can see some relations between the mechanical energy and the concepts described earlier: V ( x) = 1 2 & mx + 2 V = V1 x2 x1 0 ∫ (k x + k x )dx = 2 mx& 0 1 3 x 1 2 + V3 > V2 > V1 Fig. the level curves V ( x1 .

. using & physical meaning. this function is locally positive definite.3 (Global Stability) Assume that there exists a scalar function V of the state x.8 Asymptotic stability_______________________ Fig. composed of the sum of the potential energy and the kinetic energy.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ A Lyapunov function can be given simple geometrical interpretations. we can conclude that the origin is a stable equilibrium point. In Fig. then the stability is asymptotic. Its time derivative yields & & V (x) = θ& sin θ + θ&θ& = −θ& 2 ≤ 0 Therefore. 2 2 V ( x1 . by involving the above theorem.7 Local stability___________________________ A simple pendulum with viscous damping is described as & θ& + θ& + sin θ = 0 Consider the following scalar function 1 V (x) = (1 − cosθ ) + θ& 2 2 & its derivative V along any system trajectory is 2 2 2 2 & V = 2( x1 + x 2 )( x1 + x 2 − 2) Thus. namely that the damping term absorbs energy. & because V (x) is only negative semi-definite. is locally negative definite in the 2-dimensional ball B 2 . In fact. we must go through two steps: choosing a positive Lyapunov function. ___________________________________________________________________________________________________________ 11 Chapter 3 Fundamentals of Lyapunov Theory .9 Illustrating Definition 3. this function represents the total energy of the pendulum. Example 3. If. in the region defined by ( x1 + x2 ) < 2 . In Fig. Therefore.2 (Local stability) If. Intuitively. 3. x 2 ) = x1 + x2 Fig. we can see the reason why V (x) ≤ 0 .e. actually. 2 2 i. x2 ) is seen always point down an inverted cup. the state point is seen to move across contour curves corresponding to lower and lower value of V . there exists a scalar function V (x) with continuous first partial derivatives such that • V (x) is positive definite (locally in B R0 ) & • V (x) is negative semi-definite (locally in B R0 ) then the equilibrium point 0 is stable. the point denoting the value of V ( x1 .10 Illustrating Definition 3.e. Actually.8 for n=2 using contour curves 3.. 3. & V (x) is precisely the power dissipated in the pendulum. However. this condition indicates that − c(x ) ’pushes’ the system back towards its rest position x = 0 . V Obviously.4.9 A class of first-order systems_______________ Consider the nonlinear system & x + c( x) = 0 where c is any continuous function of the same sign as its scalar argument x . __________________________________________________________________________________________ V 0 x2 x(t ) x1 Example 3. we cannot draw conclusion on the asymptotic stability of the system. with continuous first order derivatives such that • V (x) is positive definite & • V (x) is negative definite • V (x) → ∞ as x → ∞ then the equilibrium at the origin is globally asymptotically stable. As a mater of fact. the & derivative V (x) is locally negative definite in B R0 . 3.10. but is otherwise arbitrary. In applying the above theorem for analysis of a nonlinear system.8 for n=2 V = V2 V = V1 x1 0 V = V3 V3 > V2 > V1 Let us study the stability of the nonlinear system defined by 2 2 2 & x1 = x1 ( x1 + x 2 − 2) − 4 x1 x 2 2 2 2 & x2 = 4 x1 x2 + x 2 ( x1 + x2 − 2) x2 around its equilibrium point at the origin. and then determining its derivative along the path of the nonlinear systems. __________________________________________________________________________________________ Lyapunov theorem for global stability Theorem 3.9. in a ball B R0 .Applied Nonlinear Control Nguyen Tan Tien . the above theorem indicates that the origin is asymptotically stable.2002. i. such as x c( x) > 0 ∀x ≠ 0 .2 Equilibrium point theorems Lyapunov’s theorem for local stability Theorem 3. with this Lyapunov function. Example 3. 3.

Theorem 3. the origin is a globally asymptotically stable equilibrium point. the & system x = − x 3 is globally asymptotically convergent to x = 0 . Therefore. c (x ) Local invariant set theorem The invariant set theorem reflect the intuition that the decrease & of a Lyapunov function V has to graduate vanish (i.e. where a trajectory starting from within the bounded region Ω l is seen to converge to the largest invariant set M. ___________________________________________________________________________________________________________ 12 Chapter 3 Fundamentals of Lyapunov Theory . if the set R is itself invariant (i.15). . __________________________________________________________________________________________ M be the largest invariant set in R. if once & V = 0 .10______________________________________ Consider the nonlinear system 2 2 & x1 = x2 − x1 ( x1 + x2 ) 2 2 & x2 = − x1 − x 2 ( x1 + x2 ) The origin of the state-space is an equilibrium point for this 2 2 system. sin 2 x ≤ sin x ≤ x .13). since for x ≠ 0 . Example 3 . and let V (x) be a scalar function with continuous first partial derivatives. If for a particular choice of & Lyapunov function candidate V .14.11______________________________________ Asymptotic stability of the mass-damper-spring system For the system (3. it also implies that c(0) = 0 (Fig. Note that the set R is not necessarily connected. since it tends to infinity as x → ∞ . The function V is radially unbounded.For a given system. Using the invariant set theorem. 3. where the set M consists only of the origin. V =l V Example 3 . we can show that the system is actually asymptotically stable. the theorems in Lyapunov analysis are all sufficiency theorems. Let V be the positive definite function V = x1 + x2 .2). The asymptotic stability result in the local Lyapunov theorem can be viewed a special case of the above invariant set theorem.13 The function c(x ) & For instance. 3. we cannot draw any conclusions on the stability or instability of the system – the only conclusion we should draw is that a different Lyapunov function candidate should be tried. then ≡ 0 for all future time). with f continuous. equilibrium points or limit cycles) within R . then M=R The geometrical meaning of the theorem is illustrated in Fig..3 Invariant set theorem Definition 3. Assume that • for some l > 0 . and l 0 x Fig. 3.M is the union of all invariant sets (e. TO do this..4 (Local Invariant Set Theorem) Consider an autonomous system of the form (3. . Then.Along the same line. Similarly. the region Ω l defined by V (x) < l is bounded & • V (x) ≤ 0 for all x in Ω l & Let R be the set of all points within Ω where V (x) = 0 .3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Since c is continuous. ) V has to converge to zero) because V is lower bounded. however. ⊗ Note that: . x1 Fig. the system x = sin 2 x − x is globally convergent to x = 0 .. 2 2 & Its derivative along any system trajectory is V = −2( x1 + x 2 ) 2 which is negative definite. 3.e. even about local stability.13).2002.14 Convergence to the largest invariant set M Let us illustrate applications of the invariant set theorem using some examples.In particular.Applied Nonlinear Control Nguyen Tan Tien . nor is the set M. specific choices of Lyapunov functions may yield more precise results than others. & Notice that while this system’s linear approximation ( x ≈ 0) is inconclusive. we only have to show that the set M contains only one point. the condition on V are not met. the actual nonlinear system enjoys a strong stability property (global asymptotic stability). A precise statement of this result is as follows. Its derivative & & & is V = 2 x x = −2 x c ( x) .14) in the local & equilibrium point theorem.Many Lyapunov function may exist for the same system. Thus V < 0 as long as x ≠ 0 . 3. because V is only negative semidefinite according to (3. Note that the globalness of this stability result also implies that the origin is the only equilibrium point of the system. we can only draw conclusion of marginal stability using the energy function (3. so that x = 0 is a globally asymptotically stable equilibrium point.9 A set G is an invariant set for a dynamic system if every system trajectory which starts from a point in G remains in G for all future time. Consider as the Lyapunov function candidate the square of distance to the origin V = x 2 . __________________________________________________________________________________________ Ωl R M x0 x2 ⊗ Note that: . every solution x(t ) originating in Ω l tends to M as t → ∞ .g.4.

__________________________________________________________________________________________ Fig.. ⊗ Note that: .11 actually represents a very common application of the invariant set theorem: conclude asymptotic stability of & an equilibrium point for systems with negative semi-definite V . by enlarging the involved region to be the whole space and requiring the radial unboundedness of the scalar function V . actually converges to the limit cycle. Actually.2) other than the trivial trajectory x ≡ 0 Then. x) . 3. 3. For any arbitrary positive number l . Since both the limit circle and the origin are invariant sets. along which the state vector moves clockwise. the 2 2 region Ω l .The largest connected region of the form Ω l within Ω is a domain of attraction of the equilibrium point. the above theorem does not guarantee that Ω is invariant: some trajectories starting in Ω but outside of the largest Ω l may actually end up outside Ω . combined with a third condition on the trajectories within R. is bounded. Its derivative 10 6 4 2 & V = −8( x1 + 3x 2 )( x1 + 2 x2 − 10) 2 4 2 & Thus V is strictly negative. Assume that in a certain neighborhood Ω of the origin • is locally positive definite & • V (x) is negative semi-definite & • the set R defined by V (x) = 0 contains no trajectories of (3.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ & The set R defined by x = 0 . Furthermore. but not necessarily the whole domain of attraction. The following corollary of the invariant set theorem is more specifically tailored to such applications. is bounded. the largest invariant set M in R then contains only the equilibrium point 0. because the function V is not unique. a contradiction to the definition. This implies that the x trajectory will immediately move out of the set R and thus also out of the set M. The first equation is simply that defining the limit cycle.The above corollary replaces the negative definiteness & condition on V in Lyapunov’s local asymptotic stability & theorem by a negative semi-definiteness condition on V . since d 4 2 10 6 4 2 ( x1 + 2 x 2 − 10) = −(4 x1 + 12 x 2 )( x1 + 2 x2 − 10) dt which is zero on the set. The set R is simply the origin 0. any trajectory starting within the circle converges to the origin.15 Convergence to a limit circle Moreover.Applied Nonlinear Control Nguyen Tan Tien . we see that the invariant set actually represents a limit circle. which is an invariant set (since it is an equilibrium point). a domain of attraction is explicitly determined by the invariant set theorem. For l = 1 . All the conditions of the local invariant set theorem are satisfied and. Let us show that the largest invariant set M in this set R contains only the origin. x 2 ) = x1 + x2 < 1 .13 Attractive limit cycle_____________________ Consider again the system 7 4 2 & x1 = x2 − x1 ( x1 + 2 x 2 − 10) Example 3. Is this limit circle actually attractive ? Let us define a Luapunov 4 2 function candidate V = ( x1 + 2 x2 − 10) 2 which represents a measure of the “distance” to the limit circle. all system trajectories starting in Ω l converge either to the limit cycle or the origin (Fig. the collection of states with zero velocity.The set Ω itself is not necessarily a domain of attraction. the acceleration at that point is && = −(k 0 / m) x − (k1 / m) x 3 ≠ 0 . i. defined by V ( x1 . Corollary: Consider the autonomous system (3. then. which surrounds the limit circle.2).15) ___________________________________________________________________________________________________________ 13 Chapter 3 Fundamentals of Lyapunov Theory . while the second equation is verified only at the origin.2002. Assume that M contains a point with a non-zero position x1 . excluding the origin. The motion on this invariant set is described (equivalently) by either of the equations & x1 = x 2 3 & x 2 = − x1 Therefore. Any state trajectory starting from the region within the limit cycle. the equilibrium point at the origin can actually be shown to be unstable. therefore.e.8. Thus. . the largest connected region of the form (defined by V (x) < l ) within Ω is a domain of attraction of the equilibrium point. and let V (x) be a scalar function with continuous partial derivatives. . __________________________________________________________________________________________ Example 3 . __________________________________________________________________________________________ x2 limit cycle 0 x1 Example 3 . Indeed.12 Domain of attraction____________________ Consider again the system in Example 3. in which cases V = 0 . except if x1 + 2 x 2 = 10 or 10 6 & x1 + 3 x 2 = 0 . Global invariant set theorem The above invariant set theorem and its corollary can be simply extended to a global result. with f continuous. the region Ω l . the equilibrium point 0 is asymptotically stable. & x2 = 3 − x1 5 4 − 3 x 2 ( x1 2 + 2 x2 − 10) 4 2 Note that the set defined by x1 + 2 x2 = 10 is invariant. Thus. the set M simply consists of their union. or the whole horizontal axis in the phase & plane ( x.

3. c( x) = α x ) .14 A class of second-order nonlinear systems___ Consider a second-order system of the form && + b( x) + c( x) = 0 & x where b and c are continuous functions verifying the sign & & & & conditions x b( x) > 0 for x ≠ 0 and x c( x) > 0 for x ≠ 0 . x = 0] . The first two of these equilibrium points are stable. __________________________________________________________________________________________ Example 3 .5 System Analysis Based on Lyapunov’s Direct Method How to find a Lyapunov function for a specific problem ? There is no general way of finding Lyapunov function for ___________________________________________________________________________________________________________ 14 Chapter 3 Fundamentals of Lyapunov Theory . with the above sign conditions simply indicating that the otherwise arbitrary function b and c actually present “damping” and “spring” effects. in other words. Because of the importance of this theorem. A positive definite function for this system is x 1 2 & V = x + c( y ) dy . Now V = 0 ⇒ x = 0 or x = ±1 . x b( x) = 0 only if & = 0 . if the integral c(r )dr is unbounded as x → ∞ . since they correspond to local minima of V (note again that linearization is inconclusive about their stability). Its derivative V = − x 2 − 1 x 4 . Then all solutions globally asymptotically converge to M as t → ∞ For instance. 0 ∫ x then V is a radially unbounded function and the equilibrium point at the origin is globally asymptotically stable. x = 0 . Together with the continuity assumptions.17).e.5 (Global Invariant Set Theorem) Consider an autonomous system of the form (3. any small deviation in the x direction will drive the trajectory away from it.17 The functions b(x ) and c(x ) Furthermore. 2  & This function has two minima. The dynamics of a mass-damper-spring system with nonlinear damper and spring can be described by the equation of this form. let us present an additional (and very useful) example. Let us consider each of cases: πx & x=0 ⇒ && = sin x − x ≠ 0 except if x = 0 or x = ±1 2 x = ±1 ⇒ && = 0 x Thus the invariant set theorem indicates that the system & converges globally to or ( x = −1. namely [ x = 0. or simply by noticing that because that point x is a local maximum of V along the x axis. A nonlinear R-L-C (resistorinductor-capacitor) electrical circuit can also be represented by the above dynamic equation (Fig. 3.16) vC = c (x ) v L = && x & v R = b(x ) 0 & x 0 x & Fig. Furthermore. according to the global invariant set theorem. and M be the largest invariant set in R. x = 0) . at x = ±1. the virtual Chose the Lyapunov function V = ∫  y − sin 0 x Fig. the above sign conditions are simply the necessary and sufficient conditions for the system’s stability (since they are equivalent to the conditions α1 > 0. By contrast.2). the largest invariant set M in R & contains only one point. the equilibrium point & ( x = 0. the sign conditions b and c are simply that b(0) = 0 and c = 0 (Fig. 3. with R & being the set defined by x = 0 . Now x = 0 implies that && = −c (x) . Assume that & • V (x) ≤ 0 over the whole state space • V (x) → ∞ as x → ∞ as long as x ≠ 0 .3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Theorem 3.16 A nonlinear R-L-C circuit Note that if the function b and c are actually linear & & (b( x) = α1 x. as can be shown from linearization ( && = (π / 2 − 1) x) . Differentiating V .Applied Nonlinear Control Nguyen Tan Tien . we obtain & & power “dissipated” by the system.. Example 3 . and a local maximum in x (a saddle point in the state-space) at & & & x = 0.13 is actually global: all system trajectories converge to the limit cycle (unless they start exactly at the origin. which is an unstable equilibrium point). 3. __________________________________________________________________________________________ ∫ & &x & & & & & & & V = x && + c( x) x = − x b( x) − x c( x) + c( x) x = − x b( x) ≤ 0 which can be thought of as representing the power dissipated & & in the system.α 0 > 0 ). Use of the local invariant set theorem indicates that the origin is a locally asymptotically stable point. 3. by hypothesis. x = 0 . & b(x) c (x ) & Let R be the set of all points where V (x) = 0 . Thus the system cannot get “stuck” at an equilibrium value other than x = 0 . which can be thought of as the sum of 2 0 the kinetic and potential energy of the system. the above theorem shows that the limit cycle convergence in Example 3.2002.15 Multimodal Lyapunov Function___________ Consider the system && + x 2 − 1 x 3 + x = sin & x πx 2 1 2 & x + 2 π y  dy . i. and let V (x) be a scalar function with continuous first partial derivatives. which is non-zero & x x ⊗ Note that: Several Lyapunov function may exist for a given system and therefore several associated invariant sets may be derived. with f continuous. x = 0) is unstable.

a square n × n matrix M is said to be positive semi-definite (p.A necessary and sufficient condition for a symmetric matrix M to be p.Any square n × n matrix can be represented as the sum of a symmetric and a skew-symmetric matrix. skew-symmetric.d. For instance.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ nonlinear system.d.37) where U T U = I . And the global asymptotical stability is guaranteed.10 A square matrix M is symmetric if M=MT (in other words. Since x M x is a scalar.e.5.19) is so-called Lyapunov equation. matrix is invertible.e. matrix Q. i.18 ______________________________________ Consider again the second order linear system in Example p  p 3. __________________________________________________________________________________________ ⊗ Note that: . p 21 p 22   where due to the symmetry of P.18. where P is a given symmetric positive definite matrix.d. let us consider a & quadratic Lyapunov function candidate V = xT P x .A . and therefore the linear system is 1 1   globally asymptotically stable.d. is that all its eigenvalues be strictly positive. M (t ) ≥ α I . Example 3 .5.There are some following facts • λmin (M ) x 2 ≤ xT Mx ≤ λmax (M ) x 2 • x T Mx = xT U T ΛUx = z T Λz where Ux = z • λmin (M ) I ≤ Λ ≤ λmax (M ) I • zT z = x 2 The concepts of positive semi-definite. Then the Lyapunov equation is  p11 p12   0 4  0 − 8   p11 p12  − 1 0   p 21 p 22  − 8 − 12 + 4 − 12  p 21 p 22  =  0 − 1         whose solution is p11 = 5 .1 Lyapunov analysis of linear time-invariant systems Symmetric. Therefore. • choose a positive definite matrix Q • solve for P from the Lyapunov equation • check whether P id p. is that its diagonal elements be strictly positive.d. x M x = − x M x which yields ∀x. If P is p. __________________________________________________________________________________________ M = U T ΛU (3. the unique matrix P solution of the Lyapunov equation (3. 3.(3. A square matrix M is skewsymmetric if M = −M T (i.s. In this section.A necessary condition for a square matrix M to be p. Its derivative yields & & & V = x T P x + x T P x = -xT Q x where A T P + P A = -Q (3. and negative semi-definite can be defined similarly.. Note that Q may be not p.17 ______________________________________ 4 .d. then . Lyapunov functions for linear time-invariant systems & Given a linear system of the form x = A x . p12 = p 22 = 1 . . . . Theorem 3.d.A p.) if x ≠ 0 ⇒ xT M x > 0 .. Definition 3. . ⊗ Note that: . matrix Q.d. j M ij = − M ji ).. Consider the second order linear system with A =  0   − 8 − 12 If we take P = I . This can be shown in the following decomposition M + MT M . to derive a p.) if ∀x. Λ is a diagonal matrix containing the eigenvalues of M . instead.19) be symmetric positive definite.The quadratic function associated with a skew-symmetric matrix is always zero..d.d.Applied Nonlinear Control Nguyen Tan Tien .2002. we have to use experience.6 A necessary and sufficient condition for a LTI & system x = A x to be strictly stable is that. The corresponding matrix P = 5 1 is p.d.16) In the designing some tracking control systems for robot. The − 4 − 24   matrix Q is not p. for any symmetric p. Let us take Q = I and denote P by P =  11 12  . no conclusion can be draw from the Lyapunov function on whether the system is stable or not. x T M x = 0 (3. ∀t ≥ 0.d.19) (3. even for stable systems.. and positive definite matrices Definition 3. ∀i.16) is a necessary and sufficient condition for a matrix M to be skew-symmetric. then (1 / 2)x T P x is a Lyapunov function for the linear system. we discuss a number of techniques which can facilitate the otherwise blind of Lyapunov functions.d. j M ij = M ji ). The definition of skew-symmetric matrix implies that xT M x = − xT M T x .d. 3. matrix M can always be decomposed as T T T A more useful way of studying a given linear system using quadratic functions is. this fact is very useful because it can simplify the control law.18) (3. negative definite. Let M be a n × n skew-symmetric matrix and x is an arbitrary n × 1 vector.11 A square matrix M is positive definite (p. if ∀i.Q = P A + A T P =  0 −4  . A time-varying matrix M(t) is uniformly positive definite if ∃α > 0. x T M x ≥ 0 . and physical insights to search for an appropriate Lyapunov function. Faced with specific systems. intuition.2 Krasovskii’s method Krasovskii’s method suggests a simplest form of Lyapunov function candidate for autonomous nonlinear systems of the ___________________________________________________________________________________________________________ 15 Chapter 3 Fundamentals of Lyapunov Theory . matrix P from a given p. Example 3 .MT M= + 1 24 42 3 1 24 42 3 symmetric skew− symmetric . p 21 = p12 .

Indeed. implies that the state x converges to the origin with a rate of at least γ / 2 . Since matrix theory shows that P ≤ λmax (P ) I and λmin (Q) I ≤ Q . the smallest eigenvalue of the matrix Q by λmin (Q) .P0 ) + (P . __________________________________________________________________________________________ γ = λmin (Q) / λmax (P ) corresponding to Q = I the larger than (or equal to) that corresponding to Q = Q1 .I) is positive semi-definite. The p. a12 = x 2 . (3.26) 2 2 2 & x2 = 4 x1 x2 + x 2 ( x1 + x2 − 2) Choose the Lyapunov function candidate V = x 2 . let P0 be the solution of the Lyapunov equation corresponding to Q = I is A T P0 + P0 A = −I and let P the solution corresponding to some other choice of Q A T P + PA = −Q1 Without loss of generality.P0 ) A = −(Q1 .18) implies that V ≤ −γ V . and their ratio λmax (P) / λmin (Q) by γ . and therefore. and hence the above equation implies that (P . the asymptotic stability is guaranteed.8 2 2 2 & x1 = x1 ( x1 + x 2 − 2) − 4 x1 x 2 3.23) λmax (P ) ≥ λmax (P0 ) Since λmin (Q1 ) = 1 = λmin (I ) . if W is a non-negative function. Example 3 . 2 a 21 = 3 x2 . for & nonlinear systems.3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Example 3 . (3. Then W (t ) ≤ W (0) e −α t The above Lemma implies that. V and V are not necessarily quadratic function of the states.23) represents another Lyapunov function for the system.2002. function V ( x) = 2 x1 3 2 3 + x2 + x1 x 2 2 2 (3.d.d. we obtain the p. the satisfaction of (3. This together with the fact xT P x ≥ λmin (P) x(t ) 2 . the convergence rate estimate 2 2 2 2 2 & whose derivative is V = −2 x1 − 6 x 2 − 2 x 2 ( x1 x 2 − 3 x1 x2 ) . Subtract the above two equations yields A T (P .26) guarantees the exponential convergence of W to zero.22 ______________________________________ Consider again the system in Example 3. That is dV = −2dt . The difference lies in that.. means that x T Q x ≤ V (0) e −γ t . we can assume that λmin (Q1 ) = 1 since rescaling Q1 will rescale P by the same factor.22) This is indeed p. Estimating convergence rates for nonlinear systems The estimation convergence rate for nonlinear systems also & involves manipulating the expression of V so as to obtain an explicit estimate of V . and therefore. The V (1 − V ) solution of this equation is easily found to be ___________________________________________________________________________________________________________ 17 Chapter 3 Fundamentals of Lyapunov Theory .This. the matrix (Q1 . 2 If the coefficients are chosen to be a11 = 1. we have We assume that the gradient of the undetermined Lyapunov function has the following form ∇V1 = a11 x1 + a12 x 2 ∇V2 = a 21 x1 + a 22 x2 The curl equation is ∂∇V1 ∂∇V2 = ∂x2 ∂x1 ⇒ a12 + x 2 ∂a12 ∂a = a 21 + x1 21 ∂x 2 ∂x1 λmin (Q) T x [ λmax (P ) I ] x ≥ γ V λmax (P ) & This and (3. & We can verify that V is a locally negative definite function (noting that the quadratic terms are dominant near the origin). The convergence rate estimate is largest for Q = I . a 22 = 3 .I ) Now since λmin (Q1 ) = 1 = λmax (I ) . according to xT Q x ≥ lemma.5 Performance analysis Lyapunov analysis can be used to determine the convergence rates of linear and nonlinear systems. and therefore will not affect the value of the corresponding γ .P0 ) is positive semi-definite. its & derivative is V = 2V (V − 1) .d. ∇V = x then V can be computed 1 1 2 2 as V ( x) = ∫ x1 0 x1 dx1 + ∫ x2 0 x2 dx 2 = 2 2 x1 + x 2 2 (3.20 ______________________________________ Let us use the variable gradient method top find a Lyapunov function for the nonlinear system & x1 = −2x1 2 & x2 = −2 x 2 + 2 x1 x2 Estimating convergence rates for linear system Let denote the largest eigenvalue of the matrix P by λmax (P ) . A simple convergence lemma Lemma: If a real function W (t ) satisfies the inequality & W (t ) + α W (t ) ≤ 0 where α is a real number.5.Applied Nonlinear Control Nguyen Tan Tien . a12 = a 21 = 0 & which leads to ∇V = x .4 Physically motivated Lyapunov functions 3. of P and Q implies that these scalars are all strictly positive. Therefore If the coefficients are chosen to be a11 = a 22 = 1.5.

with a rate of 1.2002.e. or “explosion”). i. However. and both have a trial and error flavor: • Hypothesize one form of control law and then finding a Lyapunov function to justify the choice • Hypothesize a Lyapunov function candidate and then finding a control law to make this candidate a real Lyapunov function Example 3 .23 Regulator design_______________________ Consider the problem of stabilizing the system && − x 3 + x 2 = u . where α = 1 − V (0) 1 + α e −2dt 2 If x(0) = V (0) < 1 . and V (t ) < α e −2t .3 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ V (x) = α e −2dt V ( 0) . .e. then α > 0 .. so that V (t ) and therefore x tend to infinity in a finite time (the system is said to exhibit finite escape time.. __________________________________________________________________________________________ 3. if V (0) > 1 . x & __________________________________________________________________________________________ ___________________________________________________________________________________________________________ 18 Chapter 3 Fundamentals of Lyapunov Theory . if the trajectory starts inside the unit circle. i.Applied Nonlinear Control Nguyen Tan Tien . This implies that the norm x(t ) of the state vector converges to zero exponentially. then α < 0 . if the trajectory starts outside the unit circle.6 Control Design Based on Lyapunov’s Direct Method There are basically two ways of using Lyapunov’s direct method for control design.

__________________________________________________________________________________________ & • The system x = − x /(1 + t ) 2 is stable (but asymptotically stable) & • The system x = − x /(1 + t ) is asymptotically stable & • The system x = −t x is exponentially stable & Another interesting example is the system x(t ) = − The solution can be expressed as x(t ) = x(t 0 ) e Since − t not x 1 + sin x 2 x ∫t0 1+sin x2 (r ) dr Extensions of the previous stability concepts Definition 4. ∀t ≥ t 0 . positive constant. It is asymptotically stable if a (r )dr = +∞ . the system is exponentially 2 convergent with rate 1 / 2 . t ) ≡ 0 ∀t ≥ t 0 (4. such problem of forced motion analysis are more appropriately treated by other methods. 4. of the form & x = f ( x. Advanced Stability Theory The objective of this chapter is to present stability analysis for non-autonomous systems.2 A first-order linear time-varying system_______ has a unique equilibrium point at the origin 0 unless A(t) is always singular. However. In practice. Non-autonomous systems ball of sufficiently small radius r .Applied Nonlinear Control Nguyen Tan Tien .1 The equilibrium point 0 is stable at t 0 if for any R > 0 . the equilibrium point 0 is unstable.3 The equilibrium point 0 is exponentially stable if there exist two positive numbers. there exists a positive scalar r ( R. 0 ∫ ∞ It is exponentially stable if there exists a strictly positive number T such that ∀t ≥ 0 .4) & Consider the first-order system x(t ) = −a(t ) x(t ) . the system & x=− a (t ) x 1+ x2 + b(t ) (4.2002. with uniform properties have some desirable ability to ___________________________________________________________________________________________________________ 19 Chapter 4 Advanced Stability Theory . Definition 4. does not have an equilibrium point. implying that the system should be able to stay at the point x* all the time.2 The equilibrium point 0 is asymptotically stable at t 0 if • it is stable • ∃r (t 0 ) > 0 such that t →∞ The asymptotic stability requires that there exists an attractive region for every initial time t 0 .1 Concepts of Stability for Non-Autonomous Systems Equilibrium points and invariant sets For non-autonomous systems. It can be regarded as a system under external input or disturbance b(t ) .4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 4. __________________________________________________________________________________________ Uniformity in stability concepts The previous concepts of Lyapunov stability and asymptotic x(t ) < R ∀t ≥ t 0 (4. Since Lyapunov theory is mainly developed for the stability of nonlinear systems with respect to initial conditions.5) ∫ a(r )dr ≥ γ . This The definition means that we can keep the state in ball of motivates us to consider the definitions of uniform stability arbitrarily small radius R by starting the state trajectory in a and uniform asymptotic stability. t ) equilibrium points x* are defined by f (x* . Example 4. with γ t t +T being a with b(t ) ≠ 0 . it is usually Otherwise.1________________________________________ The system & x=− a (t ) x 1+ x2 (4. desirable for the systems to have a certain uniformity in its behavior regardless of when the operation starts. α and λ . such as those in section 4. For instance has an equilibrium point at x = 0 .6) x (t 0 ) < r ⇒ stability for non-autonomous systems both indicate the importance effect of initial time. Thus system is stable if a (t ) ≥ 0. we can easily see that the linear time-varying system & x = A(t ) x (4. For instance. t 0 ) such that ∫ 1 + sin x (r ) dr ≥ t0 2 t x t − t0 . Its solution is x(t ) = x(t 0 ) e − ∫t0 a (r )dr t . x(t ) → 0 as t → ∞ Example 4.2) (4.9.4 The equilibrium point 0 is global stable ∀x(t 0 ) . such that for sufficiently small x(t 0 ) .3) Definition 4.1) Definition 4. x(t ) ≤ α x 0 e −λ (t −t0 ) ∀t ≥ t 0 x(t 0 ) < r (t 0 ) ⇒ x(t ) → 0 as Note that this equation must be satisfied ∀t ≥ t 0 .

The converse (ña o ñe ) is generally not û à true.1 Lyapunov’sdirect method for non-autonomous systems The basic idea of the direct method. if r = r (R) . This system has 1+ t 1 + t0 general solution x(t ) = x(t 0 ). such that any trajectory with initial states in B R0 converges to 0 uniformly in t 0 . uniform asymptotic stability always implies asymptotic stability. Definition 4. 4.1 can be chosen independent of t 0 .. functions ___________________________________________________________________________________________________________ 20 Chapter 4 Advanced Stability Theory . Theorem 4. t ) = 0 . i. But the convergence is not uniform. furthermore 3. Definition 4.d.6 The equilibrium point at the origin is locally uniformly asymptotically stable if • it is uniformly stable • there exits a ball of attraction B R0 . we extend the Lyapunov analysis results of chapter 3 to the stability of non-autonomous systems.8 A scalar time-varying function V (x.7 A scalar time-varying function V (x. R2 ) > 0 such that.2. all the stability properties of an autonomous system are uniform. the trajectory. in a ball B R0 around the equilibrium point 0. Globally positive definite functions can be defined similarly. If the condition 2 is & strengthened by requiring that V be negative definite. V (x. t ) with continuous partial derivatives such that 1. Example 4.e. this is because a larger t 0 requires a longer time to get close to the origin. as illustrated by the following example. starting from within a ball B R1 . t ) is locally positive definite if V (0. 4. then Time-varying positive definite functions and decrescent the equilibrium point is asymptotically stable.. Intuitively. ∃T ( R1 . a scalar function V (x. t ) ∂t ∂x ∂t ∂x dt (4. This drawback will partially be compensates by a simple result in section 4. Example 4. i.5 The equilibrium point 0 is locally uniformly stable if the scalar r in Definition 4.. The behavior of autonomous systems is dependent of the initial time. Uniform stability and uniform asymptotic stability: If. t ) = (1 + sin 2 t )( x1 + x2 ) 2 2 V0 (x) = x1 + x2 2 2 V1 (x) = 2( x1 + x2 ) i.4________________________________________ Consider time-varying positive definite functions as follows 2 2 V (x. t ). The solution asymptotically 1+ t converges to zero. ∀t ≥ t 0 x(t 0 ) < R1 ⇒ x(t ) < R2 ∀t ≥ t 0 + T ( R1 . t ) dominates V0 (x) and is dominated by V1 (x) because V0 (x) ≤ V (x.2 Lyapunov Analysis of Non-Autonomous Systems In this section. __________________________________________________________________________________________ ⇒ V (x. By definition. V is positive definite & 2. V (x.7) Thus.8) Lyapunov theorem for non-autonomous system stability The main Lyapunov stability results for non-autonomous systems can be summarized by the following theorem.1 (Lyapunov theorem for non-autonomous systems) Stability: If. t ) ≤ V1 (x) . we mean that for all R1 and R2 satisfying 0 < R2 < R1 ≤ R0 . Definition 4.3________________________________________ x & Consider the first-order system x = − .4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ withstand disturbances. V is decrescent The concept of globally uniformly asymptotic stability can be defined be replacing the ball of attraction B R0 by the whole state space.2002.e. a major difference in nonautonomous systems is that the powerful La Salle’s theorems do not apply. t ) ≥ V0 (x) (4. then the origin is uniformly stable. __________________________________________________________________________________________ Given a time-varying scalar function V (x. t ) is said to be decrescent if V (0. t ) is decrescent if it is dominated by a time-invariant p. will converges into a smaller ball B R2 after a time period T which is independent of t 0 .7) In other word. t ) ≥ V1 (x) (4. and if there exits a time-variant positive definite function V1 (x) such that ∀t ≥ 0. whose radius is independent of t 0 . R2 ) Definition 4. there exits a scalar function V (x. V is negative semi-definite then the equilibrium point 0 is stable in the sense of Lyapunov. concluding the stability of nonlinear systems using scalar Lyapunov functions. By uniform convergence in terms of t 0 . t ) = 0 and there exits a time-variant positive definite function V0 (x) such that ∀t ≥ t 0 .e. its derivative along a system trajectory is ∂V ∂V d V ∂V ∂V & = + + x= f ( x. can be similarly applied to non-autonomous systems.5 called Barbalat’s lemma. function. Besides more mathematical complexity. a time-variant function is locally positive definite if it dominates a time-variant locally positive definite function.Applied Nonlinear Control Nguyen Tan Tien .

and therefore. we conclude that the system is globally exponentially stable. because it dominates the time-invariant 2 2 p.17) is asymptotically stable if the eigenvalues of the symmetric matrix A(t ) + A T (t ) (all of which are real) remain strictly in the left-half complex plane ∃λ > 0. Similarly to the case of autonomous systems. function Furthermore. Applying the above result to the first equation. ∀i.. the point 0 is globally asymptotically stable. condition 3. so does x 2 . is that the time-varying system (4. V is positive definite & and V . 4. Example 4. x1 + x1 = x 2 (0) e .19) This can be readily shown using the Lyapunov function V = x T x . the matrix A(t ) remains bounded. however. t ) is negative definite.17).17) therefore.18) can be rewritten as x2 = x 2 (0) e .5 Global asymptotic stability_________________ Consider the system defined by & x1 (t ) = − x1 (t ) − e −2t x2 (t ) & x2 (t ) = x1 (t ) − x2 (t ) Chose the Lyapunov function candidate 2 2 V (x.2002. t ) is radially unbounded are all satisfied. and (4. then V is called Lyapunov function for the nonautonomous system. λi ( A(t ) + A T (t )) ≤ −λ (4. in addition.e. and & Thus. ∫ ∞ 0 A2 (t ) dt < ∞ (i.2 Lyapunov analysis of linear time-varying systems Consider linear time-varying systems of the form & x = A(t ) x (4. λi [ A(t )] ≤ −α If. because it is dominated by the time-invariant p. ___________________________________________________________________________________________________________ 21 Chapter 4 Advanced Stability Theory ∫ ∞ 0 A T (t ) A(t ) dt < ∞ (i. the integral exists and is finite) . t ) ≤ −2 ( x1 − x1 x2 + x 2 ) = −( x1 − x2 ) 2 − x1 − x 2 5 8 & x1 = −(5 + x 2 + x3 ) x1 2 & x2 = − x 2 + 4x3 & x3 = −(2 + sin t ) x3 2 Since x3 tends to zero exponentially. V (x. Example 4. and assume that at any time t ≥ 0 . Hence. is negative semi-definite. the strengthened condition 2. 0 ≤ xT x = V (t ) ≤ V (0) e − λt and therefore x tends to zero exponentially.d. ∀i. the ball B R0 is replaced by the whole state space. the system is unstable. and condition 1.2. ∀t ≥ 0. t ) = x1 + (1 + e −2t ) x 2 A simple result.Applied Nonlinear Control Nguyen Tan Tien .d.18) Sufficient smoothness conditions on the A(t ) matrix Consider the linear system (4. then the system is globally exponentially stable.8________________________________________ Consider the system defined by This shows that 2 2 2 2 & V (x. function x1 + x 2 . The solution −t t & of (4.e.17) be stable if any time t ≥ 0 .20) is globally stable exponentially stable. 2 2 & V (x.20) where A1 is constant and Hurwitz and the time-varying matrix A 2 (t ) is such that A 2 (t ) → 0 as t → ∞ and This function is p.4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Global uniform asymptotic stability: If.21) Both eigenvalues of A(t ) equal to -1 at all times. and the condition 4.. It is also decrescent. if in a certain neighborhood of the equilibrium point. its derivative along the system trajectories. V (x. then the equilibrium point at 0 is globally uniformly asymptotically stable. the integral exists and is finite) Then the system (4. It is important to notice that the result provides a sufficient condition for any asymptotic stability.. Perturbed linear systems Consider a linear time-varying system of the form & x = [ A1 + A 2 (t )] x (4. __________________________________________________________________________________________ Since LTI systems are asymptotically stable if their eigenvalues all have negative real parts ⇒ Will the system (4. so does x3 .d. the eigenvalues of A(t ) all have negative real parts ∃α > 0. ∀t ≥ 0. since & & & & V = x T x + xT x = x T ( A(t ) + A T (t )) ≤ −λ xT x = −λ V so that ∀t ≥ 0. t ) = −2 [ x1 − x1 x2 + x 2 (1 + e −2t )] 2 x1 2 + 2x 2 . the eigenvalues of A(t ) all have negative parts ? Consider the system &  x1  − 1 e 2t   x1   x2  =  0 − 1   x2  &      (4.

. i. lim sup f h. such that η does not shrink as t → ∞. t1 ) > 0. regarding t as a parameter). .o.1) and 0 be an equilibrium point.2002.23) is not satisfied for the system & x = −x + t x2 .Unlike Lyapunov’s linearization method for autonomous system.23) is satisfied..3 The linearization method for non-autonomous systems Lyapunov’s linearization method can also be developed for non-autonomous systems.o. (x. ∞) if ∀R > 0. ⊗ Note that: . then the instability of the linearized system implies that of the original non-autonomous nonlinear system.23) 4. 4. and if (4. t − t1 < η ⇒ g (t ) − g (t1 ) < R or in other words.e. and Lyapunov’s linearization method cannot be used for such systems.A function g (t ) is said to be uniformly continuous on [0.Our late results require that the uniform convergence condition (4. Given a differentiable function f of time t . since it is usually very difficult to find Lyapunov functions with a negative definite derivative. ⊗ Note that: . g (t ) is uniformly continuous if we can always find an η which does not depend on the specific point t1 .2 If the linearized system (with condition (4. no conclusion can be draw about the stability of the original nonlinear system.1) is unstable if one or more of the eigenvalues of A 0 has a positive real part . ∀R > 0. 4. the following three facts are important to keep in mind & • f → 0 ≠> f converges & The fact that f → 0 does not imply that f (t ) has a limit as t → ∞ .5.t . ___________________________________________________________________________________________________________ 22 Chapter 4 Advanced Stability Theory . then & f (t ) → 0 as t → ∞ .5 Lyapunov-Like Analysis Using Barbalat’s Lemma Asymptotic stability analysis of non-autonomous systems is generally much harder than that of autonomous systems.4 Existence of Lyapunov Functions The for any fixed time t (i. ∀t1 ≥ 0. For example.4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 4.3 Instability Theorems 4. (x. ⊗ Note that: .2 Barbalat’s lemma Lemma 4. 4. Some non-autonomous systems may not satisfy this condition. t − t1 < η ⇒ g (t ) − g (t1 ) < R .22) Therem 4. A(t ) = A 0 . . ∀t ≥ 0..t . But in some cases A is constant. the above theorem does not relate the instability of the linearized time-varying system to that of the nonlinear system. Let us denote  ∂f  A(t ) =   ∂x    x =0 (4. ∞) if ∀t1 ≥ 0. a Taylor expansion of f leads to & x = A(t ) x + f h. (4.1) around equilibrium point 0. An important and simple result which partially remedies (kha c phuï c) this situation is é Barbalat’s lemma.23). t ) If f can be well approximated by A(t ) x for any time t . we can assert its (local) stability if its linear approximation is uniformly asymptotically stable.1 Asymptotic properties of functions and their derivatives Before discussing Barbalat’s lemma itself. as stated in the following theorem: Therem 4.24) is said to be the linearization (or linear approximation) of the nonlinear non-autonomous system (4.2. contrary to what happens for autonomous nonlinear systems. Given a non-autonomous system satisfying condition (4.2 (Barbalat) If the differentiable function f (t ) has & a finite limit as t → ∞ . i.23) satisfied) is uniformly asymptotically stable.e. (4.The linearized time-varying system must be uniformly asymptotically stable in order to use this theorem. let us clarify a few points concerning the asymptotic properties of functions and their derivatives. then the equilibrium point 0 of the original non-autonomous system is also uniformly asymptotically stable. t ) x =0 ∀t ≥ 0 (4.3 If the Jacobian matrix A(t ) is constant. • f converges The fact that f (t ) has a limit as t → ∞ does not imply that f& → 0 . & the nonlinear system x = − x + x 2 / t leads to the linearized & system x = − x . Let a non-autonomous system be described by (4. ∃η ( R ) > 0.and in particular. ∀t ≥ 0.5. & • If f is lower bounded and decreasing ( f ≤ 0) . it may lead to the satisfactory solution of many asymptotic stability problem.Applied Nonlinear Control Nguyen Tan Tien . ∃η ( R.e. When properly used for dynamic systems. particularly for non-autonomous systems. Assume that f is continuously differentiable with respect to x. then it converges to a limit. ≠> & f →0 x →0 then the system & x = A(t ) x (4.23) be satisfied. and if f is uniformly continuous.A function g (t ) is continuous on [0. If the linearized system is only asymptotically stable. t and t1 play a symmetric role in the definition of uniform continuity. For example.The Jacobian matrix A thus obtained from a nonautonomous nonlinear system is generally time-varying.

3 (Lyapunov-Like Lemma) If a scalar function V (x. in addition to being negative or zero. And therefore. we study linear systems with positive real transfer function and their properties. represented in the form h( p ) = bm p m + bm−1 p m−1 + K + b0 p n + a n−1 p n−1 + K + a0 Using Barbalat’s lemma for stability analysis To apply Barbalat’s lemma to the analysis of dynamic systems. t ) is uniformly continuous in time & then V (x. one typically uses the following immediate corollary. If the transfer function of the linear subsystem is so-called positive real. This shows && is bounded. it means that the rational function h( p ) maps every point in the closed RHP (i. and therefore. The above lemma then follows from Barbalat’s lemma. Hence. and w(t ) is a bounded continuous function. if R1 > 0 is an upper bound on the & function g . There are two important differences with Lyapunov analysis: . t ) satisfies the following conditions • V (x. which looks very much like an invariant set theorem in Lyapunov analysis: Lemma 4.13_______________________________________ Consider the closed-loop error dynamics of an adaptive control system for a first-order plant with unknown parameter & e = −e + θ w(t ) θ& = −e w(t ) where e and θ are the two states of the closed-loop dynamics. such that V∞ ≤ V (x(0). 4. The difference n − m between the order of the denominator and that of the numerator is called the relative degree of the system. t ) → 0 as t → ∞ . But the invariant set cannot be used to conclude the convergence of e . representing tracking error and parameter error. means that h( p ) always has a positive (or zero) real part when p has positive (or zero) real part. Example 4. t ) is negative semi-definite & • V (x. Geometrically. and therefore from the second & & equation that y = C x is bounded. To use Barbalat’s lemma. let us check the uniform continuity & & & of V .The derivative V must be shown to be uniformly continuous.2002. Definition 4.33) is called the positive real condition. write the system in the standard form & x = A x + Bu y = Cx Since u is bounded and the linear system is strictly stable. __________________________________________________________________________________________ ⊗ Note that: Such above analysis based on Barbalat’s lemma shall be called a Lyapunov-like analysis. This in turn implies from the first & equation that x is bounded. t1 .e.33) It is strictly positive real if h( p − ε ) is positive real for some ε > 0 Condition (4.0) (this does not require uniform continuity). including the imaginary axis) into the closed RHP of h( p) . Consider the lower bounded function V = e2 +θ 2 Its derivative is & V = 2e [−e + θ w(t )] + 2θ [−e w(t )] = −2e 2 ≤ 0 The coefficients of the numerator and denominator polynomials are assumed to be real numbers and n ≥ m . & ∃t 2 (t ≤ t 2 ≤ t1 ) such that g (t ) − g (t1 ) = g (t 2 )(t − t1 ) . because θ is only guaranteed to be bounded. that e and θ are bounded. Example 4. & .The function V can simply be a lower bounded function of x and t instead of a positive definite function. Indeed.A simple sufficient condition for a differentiable function to be uniformly continuous is that its derivative be bound. thus the state x is bounded. __________________________________________________________________________________________ This implies that V (t ) ≤ V (0) . Then the system output is uniformly continuous. 4.1 PR and SPR transfer function Consider rational transfer function of nth-order SISO linear systems. because the dynamics is non-autonomous. it is often possible and useful to decompose the system into a linear subsystem and a nonlinear subsystem. although e converges to zero.10 A transfer function h(p) is positive real if Re[h( p )] ≥ 0 for all Re[ p ] ≥ 0 (4. V the approaches a finite limiting value V∞ . we can always use η = R / R1 independently of t1 to verify the definition of uniform continuity. In this section. t ) is lower bounded & • V (x. Note that. V is uniformly continuous.Applied Nonlinear Control Nguyen Tan Tien . since w is bounded by hypothesis.6 Positive Linear Systems In the analysis and design of nonlinear systems. The derivative of V is V& = −4e (−e + θ w) . and e that V & and θ were shown above to be bounded. then it has important properties which may lead to the generation of a Lyapunov function for the whole system. This is & typically done by proving that V& is bounded. Thus the system output y is uniformly continuous.12_______________________________________ Consider a strictly stable linear system whose input is bounded. Application of Babarlat’s lemma then indicates that e → 0 as t → ∞ . Indeed. the system is not asymptotically stable. ___________________________________________________________________________________________________________ 23 Chapter 4 Advanced Stability Theory .4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ ..6. This can be seen from the finite different theorem: ∀t .

4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Example 4. the real part of h( p ) is strictly positive along the jω axis.16_______________________________________ Consider the transfer function of an integrator h( p ) = value corresponding to p = σ + jω is h( p ) = 1 . and h3 has relative degree larger than 1. __________________________________________________________________________________________ σ − jω . Example 4.e. can be stated as follows Lemma 4. distinct) and the associated residues are real and non-negative • Re[ h( jω )] ≥ 0 for any ω ≥ 0 such that jω is not a pole of h( p ) The Kalman-Yakubovich lemma If a transfer function of a system is SPR. minimum-phase.e. any asymptotically stable matrix A . and a symmetric positive definite matrix L . and only if. __________________________________________________________________________________________ Theorem 4.35) is strictly positive real if. ∀ω ≥ 0 Re[h( jω )] > 0 (4.11 A transfer function h( p ) is positive real if.5 (Meyer-Kalman-Yakubovich) Given a scalar γ ≥ 0 . the phase shift of the system in response to sinusoidal inputs is always less than 900 • h( p) has relative degree of 0 or 1 • h( p ) is strictly minimum-phase (i. Its p Obviously. • h( p ) is a stable transfer function • The poles of h( p ) on the jω axis are simple (i. Consider the rational function h( p ) = h( p ) = which shows that h4 is SPR (since it is also strictly stable).9 that h( p ) is PR but not SPR. the involved system is required to be asymptotically controllable. if the transfer function H (p ) = (where the second equality is obtained by multiplying numerator and denominator by the complex conjugate of the denominator) and thus Re[h4 ( jω )] = −ω 2 +1+ ω 2 (1 − ω 2 ) 2 + ω 2 = 1 (1 − ω 2 ) 2 + ω 2 γ + c T [ pI − A]−1 b 2 ___________________________________________________________________________________________________________ 24 Chapter 4 Advanced Stability Theory . there exist positive matrices P and Q such that A T P + PA = -Q Pb = c (4. In fact. and only if.. h2 is unstable. with λ > 0 ..2002. because h1 is non-minimum phase. condition (4. vector b and c . Thus. A modified version of the KY lemma. Lemma 4.e. We σ 2 +ω 2 can easily see from Definition 4.15 SPR and non-SPR functions______________ Consider the following systems h1 ( p ) = h3 ( p ) = p −1 p 2 + ap + b 1 p 2 + ap + b h2 ( p ) = h4 ( p ) = p −1 p2 − p +1 p +1 p2 + p +1 Theorem 4.34) The above theorem implies necessary conditions for asserting whether a given transfer function h( p) is SPR: • h( p ) is strictly stable • The Nyquist plot of h( jω ) lies entirely in the RHP. and of relative degree 1) function h4 actually SPR ? We have h4 ( jω ) = jω + 1 − ω + jω + 1 2 = ( jω + 1)(−ω 2 − jω + 1) (1 − ω 2 ) 2 + ω 2 In the KY lemma. h( p) is a strictly stable transfer function ii. which is the p+λ transfer function of a first-order system. h2 and h3 are not SPR. i. while SPR functions cannot.36a) (4.34) can also be checked directly on a computer. which is summarized in the celebrated Kalman-Yakubovich (KY) lemma.4 (Kalman-Yakubovich) Consider a controllable linear time-invariant system & x = A x + bu y = cT x The transfer function h( p ) = cT [ pI − A]−1 b (4.10 A transfer function h( p) is strictly positive real (SPR) if and only if i. Is the (strictly stable. h( p ) is a positive real function. Of course. __________________________________________________________________________________________ σ + λ − jω 1 = (σ + jω ) + λ (σ + λ ) 2 + ω 2 ⊗ The basic difference between PR and SPR transfer functions is that PR transfer functions may tolerate poles on the jω axis. all its zeros are in the LHP) Example 4. there is an important mathematical property associated with its state-space representation.Applied Nonlinear Control Nguyen Tan Tien .36b) The transfer function h1 .. for example by choosing ε = λ / 2 in Definition 4. relaxing the controllability condition. one can easily see that h( p ) is strictly positive real. Re[ h( p )] ≥ 0 if σ ≥ 0 . Equivalently. Corresponding to the complex variable p = σ + jω .14 A strictly positive real function_____________ 1 .9.

10 Existence and Unicity of Solutions ___________________________________________________________________________________________________________ 25 Chapter 4 Advanced Stability Theory .Applied Nonlinear Control Nguyen Tan Tien .6. a vector q . • the involved system now has the output equation γ u 2 • the system is only required to be stabilizable (but not necessary controllable) y = cT x + 4.4 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ is SPR. H ( p ) is SPR if H ( p − ε ) is PR for some ε > 0 . Such generation is useful for the analysis and design of MIMO systems.3 Positive real transfer matrices The concept of positive real transfer function can be generalized to rational positive real matrices.4 in two aspects.8 Absolute Stability 4. and a symmetric positive definite matrix P such that A T P + P A = -q q T − ε L Pb = c + γ q This lemma is different from Lemma 4.11 An m × m matrix H ( p ) is call PR if • H ( p ) has elements which are analytic for Re( p ) > 0 • H ( p ) + H T ( p*) is positive semi-definite for Re( p ) > 0 where the asterisk * denote the complex conjugate transpose.2002.7 The Passivity Formalism 4. Definition 4. 4.9 Establishing Boundedness of Signal 4. then there exist a scalar ε > 0 .

1 Intuitive Concepts This section describes the basic concepts of feedback linearization intuitively. or controllability canonical form.5) ~ still yield h (t ) → 0 when t → ∞ . lead to exponentially stable dynamics x ( n ) + k n −1 x ( n −1) + K + k 0 x = 0 which implies that x(t ) → 0 .4) If the desired level is a known time-varying function hd (t). using simple examples.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 6. Feedback Linearization Feedback linearization is an approach to nonlinear control design. A(h) is the cross section of the tank. the ~ & equivalent input v can be chosen as v = h (t ) − α h so as to d (6. Thus.3). Feedback linearization can be applied to a class of nonlinear system described by the so-called companion form. using the control input of the form u = (v − f ) / b (6. . The dynamics (6. u ~ This implies that h (t ) → 0 as t → ∞ . so that the linear control theory can be applied.Feedback linearization technique can be view as ways of transforming original system models into equivalent models of a simpler form. the & control law v = − k 0 x − k1 x − K − k n −1 x ( n −1) with the ki chosen so that the polynomial p n + k n −1 p n −1 + K + k 0 has its roots strictly in the left-half complex plane.1 Fluid level control in a tank The dynamic model of the tank is h  d  A(h)dh  = u (t ) − a 2 gh  dt  o  & x2  x1    &    x2   x3   =  M    M     & n   f ( x ) + b( x ) u  x     where (6. the control law & v = x d ( n) − k 0 e − k1e − K − k n −1e ( n −1) (6. 6. 6. The control input is the flow u into the tank and the initial value is h0.This differs entirely from conventional linearization (such as Jacobian linearization) in that the feedback. Now.Applied Nonlinear Control Nguyen Tan Tien . From (6.1) x : the state vector f ( x). 26 ___________________________________________________________________________________________________________ . For tasks involving the tracking of the desired output xd (t).1. the close loop dynamics is ~ & h +α h = 0 Chapter 6 Feedback linearization (6.4) ~ with h = h(t ) − hd is the level error.1 Feedback linearization and the canonical form Example 6. □ ⊗ The idea of feedback linearization is to cancel the nonlinearities and imposing the desired linear dynamics.7) we can cancel the nonlinearities and obtain the simple inputoutput relation (multiple-integrator form) x ( n) = v . α is a strictly positive constant.2) (6.The central idea of the approach is to algebraically transform a nonlinear system dynamics in to a fully or partly one.4) (where e(t ) = x(t ) − x d (t ) is the tracking error) leads to exponentially convergent tracking.2002. Consider the system in companion form h output flow Fig. .3) (6.2) and (6.8) with v being an “equivalent input” to be specified. rather than by linear approximations of the dynamics.1) can be rewritten as & A(h) h = u − a 2 gh If u(t) is chosen as u (t ) = a 2 gh + A(h)v (6. b( x) : nonlinear function of the state u : scalar control input For this system. . the actual input flow is determined by the nonlinear control law u (t ) = a 2 gh − A(h) α (h) Note that in the control law (6.1: Controlling the fluid level in a tank Consider the control of the level h of fluid in a tank to a specified level hd. a is the cross section of outlet pipe.5) a 2 gh : used provide the output flow A(h) α (h) : used to rise the fluid level according to the desired linear dynamics (6. the & resulting dynamics is linear h = v Choosing v as ~ v = −α h (6.6) ∫ where. 6.

1. in the familiar form z = A z + b v . 6.2: Feedback linearization of a two-link robot Consider the two-link robot as in the Fig.2 Input-State Linearization Consider the problem of design the control input u for a single-input nonlinear system of the form lc2 l1 lc1 q1. one can use the follow control law & & & & τ 1   H 11 H 12   v1  − h q 2 − h q1 − h q 2   q1   g1   =   =  &  &  +   0 τ 2   H 21 H 22  v 2   h q1  q 2   g 2  (6.Find a state transformation z = z ( x ) and an input transformation u = u( x.11) into & z1 = −2 z1 + z 2 & z 2 = −2 z1 cos z1 + cos z1 sin z1 + a u cos(2 z1 ) (6.Applied Nonlinear Control Nguyen Tan Tien .15a) (6. state the problem of the problem of transformation stabilizing the original stabilizing the new (6.12a) (6. and vise versa). one may have to use algebraic transforms to Chapter 6 Feedback linearization ___________________________________________________________________________________________________________ 27 . Example: Consider a simple second order system & x1 = −2 x1 + a x 2 + sin x1 & x 2 = − x 2 cos x1 + u cos(2 x1 ) (6.15) input (6. Consider the following state transformation z1 = x1 z 2 = a x 2 + sin x1 which transforms (6.10) where. 6.12b) h = m 2 l1l c 2 sin q 2 g1 = m1l c1 g cos q1 + m 2 g[l c 2 cos(q1 + q 2 ) + l1 cos q1 ] g 2 = m 2 l c 2 g cos(q1 + q 2 ) Control objective: to make the joint position q1 and q 2 follows desired histories q d1 (t ) and q d 2 (t ) To achieve tracking control tasks.14) where v is equivalent input to be designed (equivalent in the sense that determining v amounts to determining u.12) nonlinear dynamics dynamics (6.2002.15b) λ : a positive number ~ ~ ~ && & The tracking error satisfies the equation q + 2λq + λ 2 q = 0 and therefore converges to zeros exponentially.9) where. Now the nolinearities can be canceled by the control law of the form u= 1 (v − cos z1 sin z1 + 2 z1 cos z1 ) a cos(2 z1 ) (6.τ1 I 1. τ 2 l2 & x = f ( x.13b) (6. it is not obvious at all what controller can stabilize it in a large region. A specific difficulty is the nonlinearity in the first equation.13b) (6. m2 q 2.11b) Fig.0).0).u ) The technique of input-state linearization solves this problem into two steps: .2 A two-link robot The dynamics of a two-link robot && & & & &  H 11 H 12   q1  − h q 2 − h q1 − h q 2   q1   g1  τ 1     &&  =  &  &  +   =   H 21 H 22  q 2   h q1 0 q 2   g 2  τ 2    (6. so that the nonlinear system dynamics is transformed into an equivalent linear time& invariant dynamics. which cannot be directly cancelled by the control input u.14) (6. m 1 I2.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Example 6. ~ ~ & & && v = q d − 2λq − λ 2 q v = [v1 v 2 ]T : the equivalent input ~ q = q − qd : position tracking error The new state equations also have an equilibrium point at (0. .11a) (6. 6. ⊗ When the nonlinear dynamics is not in a controllability canonical form.2 first put the dynamics into the controllability canonical form before using the above feedback linearization design. leading to a linear input-state relation & z1 = −2 z1 & z2 = v Thus. q = [q1 q 2 ]T : joint angles τ = [τ 1 τ 2 ]T : joint inputs (torques) 2 2 2 H 11 = m1l c1 + I 1 + m 2 (l1 + l c 2 + 2l1l c 2 cos q 2 ) + I 2 2 H 12 = H 21 = m2 l1c 2 cos q 2 + m2 lc 2 + I 2 2 H 22 = m 2 l c 2 + I 2 Even though linear control design can stabilize the system in a small region around the equilibrium point (0. v ) .Use standard linear technique to design v .11) using the original using the new transformation control input u input v (6.

29) & x3 = 2 x1 +u y = x1 Apply the same control law to the second dynamic equation..The above controller does not guarantee the stability of internal dynamics. which is non-autonomous and nonlinear.18a) (6. 6. (6. If we choose the control input to be in the form u= 1 (v − f 1 ) x2 +1 (6.What classes of nonlinear systems can be transformed into linear systems ? . However. then e(t ) ≡ 0.21) represents an explicit relationship between y and u . otherwise.19a) (6.19b) (6. except at the singularity point such that x 2 = −1 .30) & & let us differentiate the output y = x1 = sin x 2 + ( x 2 + 1) x 3 . y d (t ) and its time derivatives are assumed to be known and bounded.27a) (6. one could chose k1 = 2. this control law corresponds to the original input 1 (−2 a x 2 − 2 sin x1 − cos x1 sin x1 + 2 x1 cos x1 ) a cos(2 x1 ) (6.24) The closed-loop system under the above control law is represented in the block diagram in Fig. It is linear and controllable.3. and choosing the new input v such as & v = &&d − k1e − k 2 e y (6.17) The original state x is given from z by u= x1 = z1 x 2 = ( z 2 − sin z1 ) / a (6. the nonlinearity in (6.3 Input-Ouput Linearization Consider a tracking control problem with the following system & x = f ( x. In term of the original state. 6.20a) (6. perfect tracking is achieved.Full state measurement is necessary in implementing the control law. ⊗ Note that: . For instance. Example 6.The control law is defined anywhere. and we apply a simple linear doubleintegrator relationship between the output and the new input v.21) (6.25) linearization loop pole-placement loop z z=z (x) which represents an exponentially stable error dynamics. & & x2 + x2 = yd − e (6.21) is canceled.Applied Nonlinear Control Nguyen Tan Tien .u ) y = h( x ) (6.16) && = ( x 2 + 1) u + f1 ( x ) y f1 ( x ) 5 = ( x1 2 + x 3 )( x 3 + cos x 2 ) + ( x 2 + 1) x1 (6.23) & resulting in the stable closed-loop dynamics z1 = −2 z1 + z 2 & and z 2 = −2 z 2 . 0 v=. there are two equations: .3: Internal dynamics Consider the nonlinear control system 3 &  x1   x 2 + u   &  =  x2   u    Fig. The design of tracking controller for this doubley integrator relation is simple using linear technique. leading to the internal dynamics 3 To generate a direct relationship between the output and input. . consider the new dynamics (6. i.18b) where v is the new input to be determined.2002.20b) (6. && = v .22) Clearly.e.15). We now obtain ___________________________________________________________________________________________________________ Chapter 6 Feedback linearization 28 .27b) Control objective: to make the output y (t ) track a desired trajectory y d (t ) while keeping the whole state bounded.3 Input-State Linearization ⊗ To generalize the above method.29) and y d is differentiate again.20c) (6. k 2 are positive constant. letting e = y (t ) − y d (t ) be the tracking error.28) (6. The tracking error of the closed-loop system is given by && & e + k 2 e + k1 e = 0 (6. ∀t ≥ 0 . & e+e = 0 (6.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Now. Consider the third-order system & x1 = sin x 2 + ( x 2 + 1) x 3 5 & x 2 = x1 + x 3 y = x1 Control objective: to make y track to y d (t ) 3 & & y = x1 = x 2 + u ⇒ 3 & u = − x 2 − e (t ) + y d ( t ) (6.v) & x =f(x. let us & the facts that e is guaranteed to be bound by (6. k 2 = 0 or v = −2 z 2 (6. in view of & Since y is still not directly relate to the input u .1.How to find the proper transformations for those which can ? 6. Using the well known linear state feedback control law v = − k1 z1 − k 2 z 2 .20d) yields exponential convergence of e to zero. . e(t ) converge to zero exponentially. if initially e(0) = e(0) = 0 . & Therefore.u) x where k1 .k Tz u=u (x.

4 Input-Output Linearization of SISO System ___________________________________________________________________________________________________________ Chapter 6 Feedback linearization 29 . (6.28) does represent a satisfactory tracking control law for the system (6. .27a) (6.choose u to cancel the nonlinearities and guarantee tracking convergence.2002. where D is positive constant. control design based on input-output linearization can be made in three steps: . if the zero-dynamics is globally exponentially stable only local stability is guaranteed for the internal dynamics. . for the system (6.In linear systems. and x 2 > 0 when x 2 < − D1 / 3 .30) that & & x 2 ≤ D1 / 3 .45) This zero-dynamics is easily seen to be asymptotically stable 2 by using Lyapunov function V = x 2 . since x 2 < 0 when x 2 > D1 / 3 . ⊗ To summarize. the stability of the zero-dynamics implies the global stability of the internal dynamics. the resulting internal dynamics is unstable. .In nonlinear systems. given any trajectory & y d (t ) whose derivative y d (t ) is bounded.27).study the stability of the internal dynamics. Thus we can conclude from (6.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ & assumed to be bounded. we have y d (t ) − e ≤ D .27b) 3 & & the out put y = x1 ≡ 0 → y = x1 ≡ 0 → u ≡ − x 2 .differentiate the output y until the input u appears. 6.3 Input-State Linearization of SISO Systems 6.27) 3 &  x1   x 2 + u   &  =   x2   u    y = x1 (6. .2 Mathematical Tools 6.27a) is replaced by & x 2 = −u .Applied Nonlinear Control Nguyen Tan Tien . Therefore. ⊗ Note: if the second state equation in (6. For instance. ▲ The internal dynamics of linear systems ⇒ refer the test book ▲ The zero-dynamics Definition: The zeros-dynamics is defined to be the internal dynamics of the systems when the system output is kept at zero by the input. ⊗ The reason for defining and studying the zero-dynamics is that we want to find a simpler way of determining the stability of the internal dynamics. hence the zero-dynamics is 3 & x2 + x2 = 0 (6.

2002. we get ~ ≤ Φ/ λ−λt = ε .a. bounds (7.e.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 7.Time-varying surface S (t ) in the state-space R (n) by the scalar equation s ( x. S (t ) Bounds on s can be directly translated into bounds on the tracking error vector ~ .2 The sliding condition ___________________________________________________________________________________________________________ Chapter 7 Sliding Control 30 .2) Fig. i. Condition: For the tracking task to be achievable using a finite control u .4) where ε = Φ/ λ . s (T ) dT −λ (t −T ) f ( x ) : unknown.1 Sliding Surfaces Consider the SI dynamic system By definition (7.3). i = 0.Tracking error vector ~ & x x x x ≡ x − x = ~ ~ L ~ ( n −1) d (7.A so-called sliding control methodology is introduced. 7.Applied Nonlinear Control Nguyen Tan Tien . t ) = 0 . unknown bounded but known sign = (Φ / λ )(1 − e − λt ) ≤ Φ / λ . 1st-order problem of keeping the scalar s at zero can now be achieved by choosing the control law u of (7.The nonlinear system with structured or unstructured uncertainties (model imprecision) is considered. noting that ~ (i ) ≤  Φ 1 + λ  = ( 2 λ ) i ε is  x  n −1−i  λ    λ  ~(0) ≠ 0 .1) Fig. indeed s ≡ 0 represents a linear differential equation whose unit solution is ~ ≡ 0 . n = 3 → s = && + 2 λ ~ + λ2 ~ . t ) =  + λ    dt n −1 ~ . one has z1 ≤ φ / λn −1−i . s (t ) ≤ Φ n −1 x ⇒ ∀t ≥ 0. Fig. Apply the same procedure. 7. the problem of tracking x ≡ x d is equivalent to that of remaining on the surfaces S (t ) for all t > 0 . The simplified.2. Essentially. bounded nonlinear function b ( x ) : control gain. the initial desired state must be such that x d (0) = x (0) 7. as illustrated in Fig.1.1. we have ∀t ≥ 0. n = 2 → s = ~ + λ ~.4) are obtained bounded. ~ (i ) can be thought of as obtained through the sequence of Fig.2). ~ (i ) t ≤ (2 λ ) i ε.a Computing bounds on ~ (i ) x From previous results. within a short time-constant ( n − 1) /λ .2).3) ~ & & x x x x x For example. Furthermore.1.1. Sliding Control In this chapter: . where  d s ( x.1) such that outside of S (t ) [ ] T p λ = 1− ⇒ p+λ p+λ i . Thus it constrains trajectories to point towards the surface S (t ) .1 A Notation Simplification . Assume that ~ x ( 0) = 0 .b.1. decrease along system trajectories. λ is positive constant x (7.. : scalar control input u & x= x x L x (7. n − 1 (7. given initial condition x (7. where p = ( d / dt ) is the Laplace operator. the tracking error ~ is obtained from x s through a sequence of first order low-pass filter as shown in Fig. 7. L . 7. 7. (7. ⊗ Given initial condition (7. 1 d 2 s ≤ −η s 2 dt (7. and therefore the scalar s represents x a true measure of tracking performance.a Computing bounds on ~ x Let y1 be the output of the first filter y1 = From s ≤ Φ we thus get 1 n −1blocks [ ( n −1) T ] : state vector ∫e y (t ) = Φ e ∫ 0 t 0 t −λ (t −T ) s (T ) dT . as measured by s 2 . ⇒ The problem of tracking the n-dimensional vector x d can be reduced to that of keeping the scalar quantity s at zero.5) where η is a strictly positive constant. In the case that x asymptotically. where z1 is the output of the ( n − i − 1) th filter. x Control objective: To get the state x to track a specific time( & varying state x d = x d x d L x dn −1) [ ] T in the presence of x Similarly. s 1 1 1 L p+λ p+λ p+λ 14444444444244444444443 ~ x x ( n ) = f ( x ) + b( x ) u y=x where. 7. . 7.Tracking error in the variable x ~ ≡ x−x x d . s ~ (i) z1 x 1 1 1 1 L L p+λ p+λ p+λ p+λ 14444 244444 14444 244444 4 3 4 3 n − i −1 blocks i blocks model imprecision on f ( x ) and b ( x ) .5) states that the squared “distance” to the surface.

5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Condition (7.3) and (7. 7.5 Filippov’s construction of the equivalent dynamics in sliding mode 7. ˆ Because f is unknown and replaced by its estimation f .3 Graphical interpretation of Eqs.2 Filippov’s Construction of the Equivalent Dynamics The dynamics while in sliding mode can be written as & s=0 (7.8) ˆ u = u − k sgn( s) (7. & x sliding mode exponential convergence finite -time reaching phase xd (t) slope .11) We then have & & & & && s = ~ + λ ~ = ( && − &&d ) + λ ~ = f + u − &&d + λ ~ x x x x x x x (7.λ x s=0 A Basic Example: Consider the second-order system && = −a(t ) x 2 cos 3x + u & x where.12) From (7. Fro instance. the equivalent control can be constructed as u eq = α u + + (1 − α )u − i.5 s<0 f+ s=0 feq fs>0 Fig. 7.7) & & x x To achieve s = 0 .3 Perfect Performance – At a Price Fig. The other interesting aspect of the invariant set S (t ) is that once on it. as a convex combination of the values of u on both side of the surface S (t ) .n=2 When the switching control is imperfect.λ x s=0 Geometrically. u eq . assume Fig.5). 7.6) that the estimation error is bounded by some known function & F = F ( x.5 x 2 cos 3x ⇒ F = 0.9) ˆ & & assume that f = −1.6) ~ & & & x s = ~ + λ ~ = 0 ⇒ && = −λ ~ x x x Hence. i. the control is chosen as & ˆ x ˆ u → u = − f + &&d − λ ~ x (7.6).. The system‘s behavior once on the surface is called sliding regime or sliding mode.. of course.1.14) 31 ___________________________________________________________________________________________________________ . Let f be an estimation value of f . which can be interpreted as the & continuous control law that would maintain s = 0 if the dynamics were exactly known. we have x u eq && x x x → u = − f + && = − f + ( &&d + ~ ) (7. & u eq = − f + &&d − λ ~ x x (7.5) is referred to as sliding surface. namely d  & x x x s =  + λ ~ = ~ + λ ~ dt   By solving (7.2002. we add to u a term discontinuous across the surface s = 0 & && = f + u eq = &&d − λ ~ x x x Chapter 7 Sliding Control (7. which corresponds to requiring that the system trajectories be tangent to the surface. 7.5) is called sliding condition. S (t ) verifying (7. This intuitive construction is summarized in Fig.6). : control input u y=x : scalar output of interest (7.1. there is chattering as shown in Fig. we obtain an expression for u called the equivalent control.e.2) is not exactly verified.13) And the system dynamics while in sliding mode is. namely (d / dt + λ )n −1 ~ = 0 . the surface S (t ) will be reach in a finite time smaller that s (t = 0) / η .5) is illustrated in Fig.5) guarantees x that if condition (7. Satisfying (7.10) & f = − a (t ) x 2 cos 3 x : unknown bounded nonlinear function ˆ with 1 ≤ a ≤ 2 . 7. (7. In order to have the system track x(t ) = x d (t ) . we choose control law as u = − f + &&d − λ ~ .4 Chattering as a result of imperfect control switchings.4 & x sliding mode exponential convergence finite -time reaching phase xd (t) slope .Applied Nonlinear Control Nguyen Tan Tien . The value of α can be obtained formally from (7. the system trajectories are defined by the equation of the set itself.e. 7. we define a sliding surface s = 0 according to (7. In order to stratify the sliding condition (7. ˆ u can be seen as our best estimate of the equivalent control.5). despite the ˆ uncertainty on the dynamics f . for a system of the form && = f + u .3 for n = 2 .3).5 x 2 cos 3 x . x ) as follows ˆ f − f ≤F (7. x (0) ≠ x d (0) . 7. The typical system behavior implied by satisfying sliding condition (7.

i. the integral can be defined x within a constant.e.21) where the (possibly time-varying or state-dependent) control gain b is unknown but of known bounds 0 < bmin ≤ b ≤ bmax (7. With s and u defined as before.9): s ≤ −η s as desired. m and c are not known accurately.19) & By choosing k = k ( x. because they only describe loosely the complex hydrodynamic effects that govern the vehicle’s motion. we  0x  x 0  dt    & ˆ x ˆ obtain.18). from (7. From (7. and may also depend explicitly on time.8) is replaced by && = f + b u x (7. 2 dt So that. letting k = F +η 1 d 2 we get from (7. Indeed.18) ≤β b The estimated controller is chosen as ___________________________________________________________________________________________________________ β −1 ≤ Chapter 7 Sliding Control 32 . The 0 t (7.16) & & m && + c x x = u x (7.Applied Nonlinear Control Nguyen Tan Tien .5) can be rewritten as s s ≤ −η s = −η s sgn(s ) ..1________________________________________ A simplified model of the motion of an under water vehicle can be written ( ) + η s sgn( s) system (7. ˆ & . this in turn leads to & ˆ ˆ ˆ ˆ x k ≥ b b −1 F + η b b −1 + b b −1 − 1 f − &&d + λ ~ x and thus to (7.15) formally unchanged.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ sgn = +1 where “sgn” is the sign function:  sgn = −1 if if s>0 s<0 where. we shall call β the gain margin of our ˆ design.15). and conversely”. the sliding mode can be interpreted that “if the error is negative. β = bmax / bmin .14) and (7. Bound (7. The constant can be chosen to obtain s (t = 0) = 0 regardless of x d (0) .5) is verified. It does not for higher-order systems. one can then easily show that the control law ˆ ˆ u = b −1 [u − k sgn( s )] with ˆ k ≥ β ( F + η ) + ( β − 1) u (7.12) and 1 d 2 ˆ ˆ & (7. .13).16) Since the control input enters multiplicatively in the dynamics.19) in the & expression of s leads to & ˆ ˆ ˆ ˆ & s = ( f − b b −1 f ) + (1 − b b −1 )(− &&d + λ ~ ) − b b −1k sgn( s ) x x & Condition (7. & s = ~+λ ~ x x ~ && + λ ~ = ( && − && ) + λ ~ & & & ⇒ s=x x x xd x & & & & & ⇒ m s = m && − m &&d + m λ ~ = −c x x + u − m &&d + m λ ~ x x x x x ˆ b (7. push hard enough in the positive direction. by letting & s = ~ + 2 λ ~ + λ2 x x & x x x ∫ ~dr − ~(0) − 2 λ ~(0) 0 t Gain Margin Assume now that (7.20) (7.e.8).To the first order system. 2 dt ⊗ Note that: .3). where f − f ≤ F .17) can then be written in the form where x : position of vehicle u : control input (force provided by a propeller) m : mass of the vehicle c : drag coefficient In practice. Then. Note that the control discontinuity has been increased in order to account for the uncertainty on the control gain b .3) 2 t  & d   t x x gives s =  + λ   ~dr  = ~ + 2 λ ~ + λ2 ~dr . using (7.8) now third-order relative to this variable. ˆ it is natural to choose our estimate b of gain b as the ˆ geometric mean of the above bounds b = bmin bmax . Example 7. u = − f + && − 2 λ ~ − λ2 ~ with x x ∫ ∫ d (7.. we can now guarantee that (7.from (7. x) in (7. Indeed. and (7.f and F need not depend only on x or x . Note that be replaced by x ∫ ~(r )dr can 0 t ∫ t ~ (r )dr . Since the control law will be designed to be robust to the bounded multiplicative uncertainty (7.14) to be large enough. instead of (7. Integral Control A similar result would be obtained by using integral control. Hence we have (( f − bbˆ or −1 & ˆ ˆ ˆ f ) + (1 − b b −1 )(− &&d + λ ~ ) − b b −1k sgn( s ) s x x ≤ −η s sgn( s ) ) & ˆ ˆ ˆ ˆ b b −1k s sgn( s) ≥ ( f − b b −1 f ) + (1 − b b −1 )(− &&d + λ ~ ) s x x ( ) & ˆ ˆ ˆ ˆ x x ⇒ k ≥ (b b −1 f − f ) + (b b −1 − 1)(− &&d + λ ~ ) sgn( s) + b b −1η so that k must verify & ˆ ˆ ˆ ˆ k ≥ b b −1 f − f + (b b −1 − 1)(− &&d + λ ~ ) + b b −1η x x ˆ ˆ ˆ Since f = f + ( f − f ). the control discontinuity k across the surface s = 0 increases with the extent of parametric uncertainty. s = s s = [ f − f − k sgn( s )] s = ( f − f ) s − k s . i.20).2002. formally letting x ∫ ~(r )dr be the variable of interest.14). They may more generally be functions of any measured variables external to system (7.15) satisfies the sliding condition.

Switching control with linear observer . reflecting the simpler structure of parametric & & uncertainty: intuitively.0 -0.0 -1. .0 2. sampling time dt = 0.23) is “tighter” than the general form (7. interpolate to get u .for instance. ~ i (t ) ≤ (2 λ ) i ε x i = 0.6. 7.1) sat[( ~ + 20 ~ ) / φ] x x The tracking performance with switching control law is given in Fig.1.8.5 1.001 sec.10): && = − a (t ) x 2 cos 3 x + u .0 7.0 1.Applied Nonlinear Control Nguyen Tan Tien . s(x.5 2. 7. this leads to tracking to within a guaranteed precision ε .5 2. 7. This can be achieved by smoothing out the control discontinuity in a thin boundary layer neighboring the switching surface B (t ) = {x. replacing in essentially assigns a low-pass filter structure to the local the expression of u the term sgn(s ) by s / Φ .The smoothing of control discontinuity inside B (t ) . Φ is boundary layer thickness.5 -1.b Control interpolation in the boundary layer Given the results of section 7.1 .a The boundary layer 0 0.0 2.5 0 0.6.7 and with smooth control law is given in Fig.5 0.0 1.2002. __________________________________________________________________________________________ Example 7.5) . regardless of uncertainty on m .6.5 x 2 cos 3 x + &&d − 20 ~ & & x x − (0. for a given problem.5 3.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ & ˆ ˆ x u = m ( &&d − λ ~ ) + c x x x ˆ& & and a control law satisfying the sliding condition can be derived as & ˆ ˆ x u = u − k sgn( s ) = m ( &&d − λ ~ ) + c x x − k sgn( s ) x ˆ& & where k is calculated from (7. Switching control law: ˆ u = u − k sgn( s) & & x x = 1. 7.5 2. chattering must be eliminated for the controller to perform properly.5 3.5 1.b illustrates this concept: ⊗ Note that: .0 3.5 4 Time (s) Time (s) φ bo un da ry Fig. 7. The constants are chosen as λ = 20. K .5 3. Fig.0 3.0 1. η = 0.0 3.25) where.23) (7. and ε = Φ/λ n-1 is the boundary layer width.1) sgn( ~ + 20 ~ ) Smooth control law with a thin boundary layer φ = 0. 7.2________________________________________ & Consider again the system (7.5 1. 7.0 2. and the use of artificial dither to reduce stiction effects.22) −φ u boundary layer ˆ u φ s Fig.5 4 Time (s) Time (s) Fig. & x 4 Tracking Error 4 Control input 3 2 1 0 -1 -2 -3 -4 0 0. and more generally guarantees that for all trajectories starting inside B(t = 0) ∀t ≥ 0.Out side of B (t ) . In general. t ) ≤ Φ} Φ>0 (7.a illustrates boundary layer for the case n = 2 .4 Direct Implementations of Switching Control Laws The main direct applications of the above switching controller include the control of electric motors.20) ˆ k ≥ β ( F + η ) + ( β − 1) u & ˆ x ≥ β ( F + η ) + ( β − 1) m ( &&d − λ ~ ) + c x x x ˆ& & Hence k can be chosen as follows & ˆ k = ( F + β η ) + m ( β − 1) ( &&d − λ ~ ) x x (7. 7.5 4 ε Fig.Switching control in place of dither 7.0 2.5 x 2 cos 3 x + 0.5 x 2 cos 3 x + 0.1. ___________________________________________________________________________________________________________ Chapter 7 Sliding Control 33 .0 1.7 Switched control input and tracking performance ε x Control Input 6 5 4 5 4 lay er Tracking Error (x10-3) 3 2 1 0 -1 -2 -3 -4 -5 3 2 1 0 -1 -2 -3 -4 0 0.8 Smooth control input and tracking performance __________________________________________________________________________________________ Fig.Switching control in place of pulse-width modulation . x and assume that the desired trajectory is x d = sin(π t / 2) .1 : ˆ u = u − k sat ( s / φ) & & = 1. choose the control law u as before (7.0 3.5 1.5 3. rather than apply some prepacked formula.2 Continuous Approximations of Switching Control Laws In general. it is a good idea to quickly rederive a control law satisfying the sliding condition.20).6.5 1.5 x 2 cos 3 x + &&d − 20 ~ x x & & − (0.5 2.5 0. u can compensate for c x x directly.Inside of B (t ) .1. n − 1 Note that the expression (7. 6 5 1.

They can be expressed directly in terms of the variable s as & s = − k ( x) s − ∆ f ( x) φ (7. and −φ(t ) on a single diagram as illustrated in Fig. & x ˆ Accordingly.28) in the form ___________________________________________________________________________________________________________ Chapter 7 Sliding Control 34 .31) (7. Consider again the system (7.5 x 2 cos 3 x + η − 0.The boundary layer thickness φ describes the evolution of dynamics model uncertainty with time.26) reflects the fact that the boundary layer attraction condition is more stringent during & boundary layer contraction ( φ < 0 ) and less stringent during & boundary layer expansion ( φ > 0 ). whose dynamics only depend on the desired state x d .3) − ∆ f (x d ) + O (ε ) 1storder filter (7.27) sat( y ) = y if y ≤ 1  and sat is the saturation function  sat( y ) = sgn( y ) otherwise  and can be seen graphically as in the following figure sat( y ) −1 1 y ⊗ Note that: .The s-trajectory represents a time-varying measure of the validity of the assumptions on model uncertainty. The control law is now ( )( ) ˆ where ∆ f = f − f . Conceptually. From (7.32) (7. to tune up the control law so as to achieve a trade-off between tracking precision and robustness to un-modeled dynamics.e.5 x 2 cos 3 x + η − φ & &2 where.5 x d cos 3 x + η ) . φ(t ) . in our smooth implementation the term k (x) sgn( s ) obtained from switched control law u is actually replaced by k (x) sat( s / φ) . with s + (− ∆ f (x) + O(ε ) ) (7.. which in turn provides tracking error ~ by further x low-pass filtering.5 xd cos 3 x d + η + λ φ & & = 0. the perturbations are filtered according to (7. Assume that φ(0) = η / λ with η = 0.The s-trajectory is a compact descriptor of the closedloop behavior: control activity directly depends on s .31) and (7.2002. i. we consider the system trajectories inside the boundary layer. using (7. i. uncertainty ∆ f (x d ) . & the quantity −φ is added to control discontinuity gain k (x) .32) & &2 k ( x ) = 0.3________________________________________ Consider again the system described by (7. d s≥φ ⇒ ( s − φ) ≤ −η dt d ⇒ ( s − φ) ≥ η s ≤ −φ dt Thus. and whose input are.Applied Nonlinear Control Nguyen Tan Tien .29) choice of φ ˆ b = b = 1 . ..1 .29) that the variable s (which is a measure of the algebraic distance to the surface S (t ) ) can be view as the output of the first order filter. while tracking error ~ is merely a filtered version of s x .29) φ We can see from (7.5).10): && = − a (t ) x 2 cos 3 x + u .5) be satisfy outside the boundary layer. Now.e. it must be chosen to be “small” with respect to high-frequency un-modeled dynamics (such as unmodeled structural modes or neglected time-delays). Recognizing this filter-like structure then allows us.9 Structure of the closed-loop error dynamics Control action is a function of x and x d . It is thus particularly informative to plot s (t ) . “perturbations”.30) & The additional term φ s in (7. we can now turn the boundary layer thickness φ so that (7.In order to satisfy (7. Since k and ∆ f are continuous in x . to & s = − k ( x) the first order. we now required that s ≥φ ⇒ 1 d 2 & s ≤ (φ .1): x ( n ) = f (x) + b(x) u . as long as high-frequency un-modeled dynamics are not excited.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ dynamics of the variable s .26). and can be monitored so as to well exploit the control “bandwidth” available. .29) to give s .29) also presents a first-order filter of bandwidth λ . where & k ( x) = k ( x) − φ (7. It suffices to let k (x d ) =λ φ which can be written from (7. in essence.26) s 1 ( p + λ ) n −1 definition of s ~ x Fig. 7. φ = −λ φ + (0.Boundary layer thickness φ can be made time-varying. we must actually modify condition (7.3). control law becomes u = u − k (x) sat( s / φ) . according to definition (7.27) can be rewritten as k ( x) = k ( x) − k ( x d ) + λ φ (7. 7.11b. we now need to guarantee that the distance to the boundary layer always decreases. Example 7.4) to rewrite (7. In order to maintain attractiveness of the boundary layer now that φ is allowed to vary with time. thus eliminating chattering.η ) s 2 dt (7.27) as & φ + λ φ = k (x d ) (7. Since λ is breakfrequency of filter (7.28) λ = 20 . Furthermore. Indeed. instead of simply required that (7. Thus chattering can be eliminated.

7.0 10 5 0 -5 -10 -15 -20 -25 0 1 2 3 4 5 6 b. References 15 1.2.0 3. in the sense that it guarantees the best tracking performance given the desired control bandwidth and the extent of parameter uncertainty.5 4 Time (s) Time (s) a. .31) is 35 ___________________________________________________________________________________________________________ .5 4 -5 Time (s) Time (s) Fig.0 3.11a Control input and resulting tracking performance 8 6 And the controller is & u = c x x + m ( &&d − λ ~ ) ˆ ˆ& & ˆ x x  & k ( x) = max c − c x 2 + max m − m && − λ ~ + max(m)η ˆ ˆ & xd x  φ = k ( x ) − λ φ & d  k = k ( x ) − φ &  ~+λ~ & s = x x  ˆ u = u − k sat( s / φ)  The results are given in Fig.21): m && + c x x = u .5 1.5 1.2.The arbitrary constant η (which formally.5 4 & ˆ ˆ & & k ≥ (c − c) x x + (m − m) ( &&d − λ ~ ) sgn( s ) + mη x x 0 0. The a priori bounds x on m and c are: 1 ≤ m ≤ 5 and 0. one can easily show that (7.5 (m/s2 ) In the case that β ≠ 1 .0 1. 6 5 5 4 & Condition (7.36) Example 7.5 4 -10 0 0.5 0 1 2 3 4 5 6 s λφ k (x d ) ≤ βd λφ 2 βd = k (x d ) βd −φ Time (s) k (x) = k (x) − k (x d ) + λ φ/β d with initial condition φ(0) defined as: φ(0) = β d k (x d (0)) / λ Time (s) c.The value of λ is selected based on the frequency range of un-modeled dynamics. then φ from (7. Control input S-trajectories (x10-2) k (x d ) ≥ λφ βd ⇒ ⇒ & φ + λ φ = β d k (x d ) & φ+ (7. 7.31) and (7.5 3.0 1.34) (7.0 2.5 3.5 x 2 cos 3x + &&d − λ ~ & & x x − (0. s.0 2.4________________________________________ A simplified model of the motion of an under water vehicle & & can be written (7.2002. __________________________________________________________________________________________ Desired Trajectories 3 1 0 -1 -2 -3 -4 Control Input 2 20 15 10 5 0 -5 acceleration velocity (m/s) distance (m) 0 0.5 ≤ c ≤ 1.Applied Nonlinear Control Nguyen Tan Tien .5 -1.33) (7. as describe above is designed as follows: & s = ~+λ ~ x x Chapter 7 Sliding Control ⊗ Remark: .5 1.5 2.0 2.0 1. the tracking error is consistently better (up to 4 times better) than that in Example 7. The control input.5 2.5 s −φ 4 Time (s) Fig. because varying the thickness of the boundary layer allow us to make better use of the available bandwidth.5 0 -0. The smooth control input using time-varying boundary layer.5 3.11b s-trajectories with time-varying boundary layer We see that while the maximum value of the time-varying boundary layer thickness φ is the same as that originally chosen (purposefully) as the constant value of φ in Example 7.5 1.0 3. λ = 20 . reflects the time to reach the boundary layer starting from the outside) is chosen to be small as compared to the average value of k (x d ) . Tracking error Fig.3) used to define sliding surfaces is the “best-conditioned” among linear dynamics.12 b.5 1.5 . 7. and s -trajectories are plotted in Fig. 7. so as to fully exploit our knowledge of the structure of parametric uncertainty.11.The desired trajectory x d must itself be chosen smooth enough not to excite the high frequency un-modeled dynamics.35) φ 0.5): s s ≤ −η s & ˆ ((cˆ − c) x& x& + (m − m) (&x& − λ ~ ) − k sgn(s)) s ≤ −mη s x & ˆ ˆ & & ˆ& & k s sgn( s ) ≥ ((c − c) x x + c x x + (m − m) ( && − λ ~ ) ) s + mη s x x & ˆ ˆ & & k ≥ ((c − c) x x + (m − m) ( && − λ ~ ) ) sgn( s ) + mη x x d d d & m s s ≤ − mη s Tracking Error (x10-3) 4 3 2 1 0 -1 -2 -3 -4 Control Input 3 2 1 0 -1 -2 -3 -4 0 0.0 3. η = 0.32) become (with β d = β (x d ) ) Tracking Error (x10-2) -5 2. .1 .0 2.0 1.5 2.5 x cos 3x + η − φ) sat[( ~ + 20 ~ ) / φ] &2 & & x x s = && − &&d + λ ~ x & & & & m s = −c x x + u − m ( &&d − λ ~ ) x x & ˆ ˆ& & ˆ x u → u = c x x + m ( &&d − λ ~ ) x ˆ u = u − k sgn( s) & ˆ& & ˆ x x = c x x + m ( &&d − λ ~ ) − k sgn( s) & & ˆ x & & & ˆ& & m s = −c x x + c x x + m ( &&d − λ ~ ) − k sgn( s ) − m ( &&d − λ ~ ) x x x & ˆ ˆ & & x x = (c − c) x x + (m − m) ( && − λ ~ ) − k sgn( s) d ⊗ Note that: .5 2.0 1. tracking error.5 1. Their estimate ˆ ˆ values are m = 5 and c = 1 .trajectories __________________________________________________________________________________________ (7.12 5 4 35 30 25 S-trajectories (x10-2) φ 4 2 0 -2 -4 -6 -8 0 0.An argument similar to that of the above discussion shows that the choice of dynamics (7.If the model or its bounds are so imprecise that F can only be chosen as a large constant.5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ ˆ u = u − k ( x) sat ( s / φ) & & x x = 1.0 -1. .0 2.5 3.

e. ν sampling is the sampling rate. while the augmenting gain k (x) according by the quantity u ′ ( β − 1) .e. better knowledge of f reduces k by a comparable quantity. L . poor dynamic models may lead to respectable tracking performance. i. margin gains in performance are critically dependent on control bandwidth λ : if large λ ’s are available. in certain case. Neglecting time-constants of order 1 / λ . m . classically λ ≤ λR ≈ 2π νR 3 (7. we see that the effects of parameter uncertainty on f have been “dumped” in gain k .. it may account for the fact that λ R may actually vary with the task. L . i. give system model (7. . one gets a condition of the form 1 (7. The desired control bandwidth λ is the minimum of three bounds (7.1).34) can be written where T A is the largest un-modeled time-delay (for instance in the actuators).5 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ constant and large. for instance. neglected time delays: along the same lines.e. given clean measurements. and conversely large modeling efforts produce only minor absolute improvements in tracking accuracy. j = 1.39) is particularly useful in an incremental mode.33) and (7. i = 1.3 The Modeling/Performance Trade-Offs The balance conditions (7. Conversely.2002. ii. a reasonable interpretation of this constrain is.4 Multi-Input System Consider a nonlinear multi-input system of the form x i ( ni ) = f i (x) + where u = [u 1 x = x ( n −1) 7.44) λn ε ≈ β d k d (7. Thus (7.41) although in practice this bound may be modulated by engineering judgment. 7. m ˆ ˆ Consider the control law (7.A well-designed controller should be capable of gracefully handling exceptional disturbances.Applied Nonlinear Control Nguyen Tan Tien . structural resonant modes: λ must be smaller than the frequency ν R of the lowest un-modeled structural resonant mode.39) ∑b j =1 m ij ( x) u j . having λR ≈ λ A ≈ λS ≈ λ 7..41-43).42) ___________________________________________________________________________________________________________ Chapter 7 Sliding Control 36 .40) [ u2 L u m ]T x ( n − 2) & L x ] : the control input vector : the state vector T In particular. to evaluate the effects of model simplification on tracking performance: ∆ε ≈ ∆ ( β d k d / λ n ) (7. It will be discussed in next section. condition (7. And it is not overly surprising that system performance be very sensitive to control bandwidth. λ typically limited by three factors: i. the term &x ˆ u ′ = −λ ~ should be added to the corresponding u . how large λ can be chosen ? In mechanical system.In the case that λ is time-varying. Ideally.43) λ ≤ λS ≈ ν sampling 5 where. sampling rate: with a full-period processing delay. i. . so that the term k sat( s / φ) simply equals λ s / β in the boundary layer.33)-(7.36) have practical implications in term of design/modeling/performance tradeoffs. taking notably into account the natural damping of the structural modes.19): u = b −1 [u − k sgn( s )] . disturbances of intensity higher than the predicted bounds which are used in the derivation of the control law.. iii. we have a condition for the form λ ≤ λA ≈ 1 3TA (7.5 Summary (7. Furthermore. Thus. the most effective design corresponds to matching these limitations.

3 0.5 1.8) However this is not good method because there may be considerable noise in the measurement && .6) If the update law is chosen to satisfy & ˆ m = −γ s v The derivative of Lyapunov function becomes & V = −λ m s 2 ≤ 0 (8. the simplest way of estimating m is ˆ m(t ) = u (t ) &&(t ) x (8. λ = 6.4 0.0 -0. ___________________________________________________________________________________________________________ Chapter 8 Adaptive Control 38 .5.5 A self-tuning controller A self-tuning controller is a controller which performs simultaneous identification of the unknown plant. x the acceleration may be close to zero.0 Time (s) Time (s) Fig.2 0.analyze the convergence properties of the resulting control but let us now generate the estimated mass parameter using a system.5 ∫ e (r ) dr 2 0 t (8.1 0. 8.0 1.14).6 0.11) 1.0 0.0 1. This total error minimization can potentially average out the effects of measurement noise.0 -0.1 0.5 ∫ w u dr ˆ m= ∫ w dr 0 t 2 0 t (8.Choose an adaptation law for adjusting those parameters dynamics) control laws (8.3) for generating the control input.5 1.5 1.5 3. The resulting estimating is Tracking Performance Parameter Estimation 2.5 and r (t ) = sin( 4t ) . __________________________________________________________________________________________ Fig.5 2. and.5 2..5 ∫w 0 (8.4 0.2 Self-tuning control of unknown mass________ The design of an adaptive controller usually involves the following three steps: Consider the control of a mass of Example 8. The prediction error is simply the error ˆ in fitting the known input u using the estimated parameter m . it is easily to show that s converges to zero.0 2.5 3. (8.e.3 How to Design Adaptive Controllers ? Example 8. 8.5 Tracking performance and parameter estimation for an unknown mass with reference path r (t ) = sin( 4t ) __________________________________________________________________________________________ ( ) (8.0 0. 8. the results of simulation for this example are given in Fig.0 0.11)(which can be written controller u plant y ˆ P −1m = w u dr ) leads to 0 ∫ t & ˆ m = − P (t ) w e ˆ a (8. 8.5 2. Relations between MRAC and ST methods 8. 8.4 Tracking performance and parameter estimation for an unknown mass with reference path r (t ) = 0 0.4.6 -0.5 3. To increase computational efficiency.0 1.6 2. λ1 = 10. 8. it is desirable to adopt a recursive formulation instead of repeatedly using (8. γ = 0.4 -0.choose a control law containing variable parameters the pole-placement (placing the poles of the tracking error . If actually. For x illustration. From (8.0 The function P (t ) is called the estimation gain.2002.7) For simplicity.5.13) Self-Tuning Controller (STC) r Then differentiation of Eq. we define P (t ) ≡ 1 t Tracking Performance Parameter Estimation 0. choosing the estimate in such a way that the total prediction error Using Barbalat’s lemma.5 1. (8.0 0. 0.0 2.0 0. To do this.14) estimator ˆ In implementation. the above estimate has to be recalculated at every new time instant. the unknown parameter m is slowly x time-varying. x(0) = x m (0) = 0 .0 0.0 0.8 0. x(0) = x m (0) = 0.13) and (8.0 2. λ2 = 25.2 -0. x(0) = xm (0) = 0 and r (t ) = 0 .5 0.4 and 8.0 0.0 0. estimation law.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 1 1 1 ~2 m s2 + m ≥0 2 2γ Its derivative yields V= ~~ & & & V = s m s + γ −1m m ~ ~ & ˆ = s (−λ m s − m v ) − γ −1m m ~ & ˆ = −λ m s 2 − γ −1m m + γ s v (8. The numerical values are chosen as m = 2 .5 with w = && .0 2.5 2.0 0.0 1.Applied Nonlinear Control Nguyen Tan Tien .5 1. the parameter estimate m is obtained by numerically integrating Eqs.11).1.12) dr 2 1. i. assume that the acceleration can be measured by an accelerometer. . & & ˆ m(0) = 0. A better approach is to estimate the parameter using a least-squares approach.5. furthermore.2 0. with the prediction error e defined as ˆ x e(t ) ≡ m(t ) &&(t ) − u (t ) .10) is minimal.5 3.0 2.0 1. In Fig. In Fig.8 0. The convergence of s to zero implies that of the & x position tracking error ~ and the velocity tracking error ~ . its update can be directly obtained by using d −1 P = w2 dt Time (s) Time (s) Fig.1.5 2. Let us still use .9) ( ) (8.1).

0 -0.5 2.0 3. Choice of control law Define a signal z (t ) as follows (n z (t ) = y m ) − β n−1e ( n−1) − K − β 0 e (8.33) is introduced with the intention of adaptively canceling the nonlinear term.5 3. 8.0 2.33) where the second term in (8.5 4.0 0.0 0.0 0.38) with β1 .Applied Nonlinear Control Nguyen Tan Tien .5 -1. Two reference signals are used: denoting the estimated parameter vector.0 1.5 2.0 -0.5 1.0 1.12 where ~ = a − a a ˆ ___________________________________________________________________________________________________________ Chapter 8 Adaptive Control 40 .34c) & ˆ a r = − sgn(b p ) γ e r Example 8.5 5.0 2.11 Tracking performance and parameter estimation with reference path r (t ) = 4 5 4 2.0 -1.K.5 -2.5 -1.5 5.34b) (8.0 4.5 3.3 Adaptive Control of Linear Systems with Full States Feedback Consider the nth-order linear system in the canonical form a n y ( n ) + a n−1 y ( n−1) + K + a0 y = u (8.5 1.5 -2.5 3 1.22).5 0. The tracking error e = y − y m then satisfies the closed-loop dynamics a n [e ( n) + β n−1e ( n−1) + K + β 0 e = v T (t )~ (t ) (3.5 -2.5 1.5 2.0 3. This represents a pole-placement controller which places the poles at positions specified by the coefficients β i .5 -1.5 5. 8. y .K.0 0 1 2 3 4 5 6 ˆ ar ˆ af ˆ ay ˆ ay 7 8 9 10 Time (s) Time (s) Time (s) Time (s) Fig.5 0. y ( n−1) are measurable.0 1.5 0.5 Fig.39) (8.0 2.0 -1. The objective of the control system is to make y closely track the response of a stable reference model & where the state components y.5 ˆ ar ˆ ay Tracking Performance Tracking Performance Parameter Estimation 1.0 1.34a) (8. Using the same procedure for the linear plant.5 2.10 Tracking performance and parameter estimation with reference path r (t ) = 4 sin(3t ) __________________________________________________________________________________________ Fig. 8. 8.12 Tracking performance and parameter estimation with reference path r (t ) = 4 sin(3t ) __________________________________________________________________________________________ Parameter convergence analysis ⇒ refer text book Extension to nonlinear plant The same method of adaptive control design can be used for the non-linear first-order plant describe by the differential equation & y = −a p y − c p f ( y) + b p u (8.32) 8. Instead of using (8.5 1. a m = 4 .0 4.0 Parameter Estimation 4.0 ˆ ar ˆ ay ˆ af Time (s) Time (s) Time (s) Time (s) Fig.0 -0.4 A first-order non-linear plant_______________ & Consider the control of the unstable plant y = y + y 2 + 3 u using the previous designed nonlinear adaptive controller.0 -1.0 2.0 0. The adaptation laws are & ˆ a y = − sgn(b p ) γ e y & ˆ a f = − sgn(b p ) γ e f a f ≡ c p / bp and (8. β n being positive constants chosen such that p n + β n−1 p n−1 + K + β 0 is a stable (Hurwitz) polynomial.5 -2.21) and define ~ ˆ a f ≡ a f − a f .0 1. now we use the control law coefficient vector a = [a n L a1 a0 ]T is unknown. bm = 4 . 8.5 2.5 3. The nonlinear in these dynamics is characterized by its linear parametrization in terms of the unknown constant c .0 1.36) and rearranging. but their signs are known. and y (0) = y m (0) = 0 .0 0.5 0.0 2.0 1.11 * r (t ) = 4 sin(3t ) ⇒ simulation results in Fig. γ = 2 .0 -0.36) where f is any known nonlinear function.5 3.0 3.5 ˆ ar Tracking Performance Tracking Performance Parameter Estimation 2 1 0 -1 -2 -3 -4 -5 0 1 2 3 4 5 6 7 8 9 10 1.5 -1.0 1.37) ˆ ˆ ˆ u = a y y + a f f ( y ) + ar r (8.0 0.40) a * r (t ) = 4 ⇒ simulation results in Fig.0 1.0 -1. 8.9 Tracking performance and parameter estimation with reference path r (t ) = 4 5 4 2. Adding both side of (8.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 6 5 4 3 2 1 0 0.0 0.0 0.5 4. & ˆ ˆ ˆ y = −a p y − c p f ( y ) + b p [a y y + a f f ( y ) + a r r ] ˆ ˆ ˆ = − ( a p − b p a y ) y − (c p − b p a f ) f ( y ) + b p a r r Comparing to (8.0 0.0 0 1 2 3 4 5 6 7 8 9 10 Parameter Estimation 3 3 2 1 0 -1 -2 -3 -4 -5 0 1 2 3 4 5 6 7 8 9 10 1.0 0. The numerical parameters are: a p = −1 . α n y m ( n) + α n−1 y m ( n−1) + K + α 0 y m = r (t ) with r (t ) being a bounded reference signal.5 2.2002.0 0 0.0 4.5 0.5 1. we can rewrite the plant dynamics as a n [ y ( n) − z ] = u − a n z − a n−1 y ( n−1) − K − a0 y Let us choose the control law to be ˆ ˆ ˆ ˆ u = a n z + a n−1 y ( n−1) + K + a0 y = v T (t )a(t ) with v (t ) = [z (t ) ˆ ˆ a(t ) = [a n & y n−1 L y ˆ ˆ a n−1 L a1 y ]T ˆ a0 ]T (3.0 2. b p = 3 .

13 Model-reference control system for relative degree 1 Let the controller be chosen as shown in Fig.47) where z = u /( p + bm ) . b =  M  . If we take kp these . The following one is particularly convenient for later adaptation design. n − 1. k= be α1 = b p − bm . It is well known from linear Chapter 8 Adaptive Control ___________________________________________________________________________________________________________ 41 . β 2 .43) (8. in our treatment. and k m is positive.the sign of k p is known . The objective of the design is to determine a control law.41a) (3. The relative degree r of this system is r = n − m .1. The desired performance is assumed to be described by a reference model with transfer function Wm ( p ) = k m Zm Rm (8. Therefore.1 Linear systems with relative degree one Choice of the control law To determine the appropriate control law for the adaptive controller. z is the output of a first-order filter with input u . the adaptation law & ˆ a = − Γ v bT P x & leads to V = − x T Q x ≤ 0 . j = 0.K. so that the plant output y asymptotically approaches y m . Example 8.e. k are controller parameters.4. 8. we must first know what control law can achieve perfect tracking when the plant parameters are perfect known. the high frequency response is ω essentially determined by k p .Applied Nonlinear Control Nguyen Tan Tien . We assume as follows . ~ ) = xT P x + ~ T Γ −1~ where both Γ and P are symmetric positive constant matrix. we will assume that nm − mm ≥ n − m . The reason for this term is that the plant frequency response at high frequency kp verifies W ( jω ) = n−m . β1 . β2 = parameters a m2 − a p2 kp to . i. The derivative V can be computed easily as & & V = − xT Q x + 2~ T v b T P x + 2 ~ Γ −1~ a a a Therefore. and α1 . km . with the control law being u = α1 z + where k p is called the high-frequency gain.the relative degree n − m is known .2002.e.the plant order n is known .. Many controller structures can be used for this purpose. b j (i = 0.45) and the reference model ym = k m ( p + bm ) p 2 + a m1 p + a m2 r (8.5 A controller for perfect tracking_____________ Consider Lyapunov function candidate a a a V (x. the coefficients ai .46) ym (t ) e r (t ) k Wm ( p ) u0 u1 u p + bm W p ( p) α1 β1 p + β 2 p + bm Fig. and P satisfies PA + A T P = −Q Q = QT > 0 & for a chosen Q .13.the plant is minimum phase 8.1.41b) L L O L L  1 0       0  0  0  M  . 8. and an associated adaptation law.K. m − 1) and the high frequency gain k p are all assumed to be unknown.. cT =  M       1  0  0  0  1 − β n−1       0 theory that the relative degree of the reference model has to be larger or equal to that the plant in order to allow the possibility of perfect tracking. 8.42) Consider the plant described by y= k p ( p + bp ) p + a p1 p + a p2 2 u (8. i.4 Adaptive Control of Linear Systems with Output Feedback Consider the linear time-invariant system presented bu the transfer function W ( p) = k p Z p ( p) R p ( p) = kp b0 + b1 p + K + bm−1 p m−1 + p m a0 + a1 p + K + a n−1 p n−1 + p n (8.40) in state space form & x = A x + b [(1 / a n ) v T ~ ] a e = cx where 1 0  0  0 1  0 A= M M M  0 0  0 − β  0 − β1 − β 2 (3.44) β1 p + β 2 y+kr p + bm (8. the transfer kp a m1 − a p1 β1 = function from the reference input r to the plant output is Wry = k m ( p + bm ) p 2 + a m1 p + a m2 = Wm ( p ) where Z m and Rm are monic Hurwitz polynomials of degrees nm and mm .6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Choice of adaptation law Rewrite the closed-loop system (3. In our adaptive control problem.

8. θ1 . As a result of the above three parts.50) The above controller in Fig. 8.47) is composed of three parts: .. * . the closed-loop system has the desired transfer function.The block for generating the (n − 1) × 1 vector ω 2 has the & same dynamics but with y as input. Chapter 8 Adaptive Control ___________________________________________________________________________________________________________ 42 .e. perfect tracking is achieved with this control law. The output here must be y (t ) = Wm ( p )r + Wm ( p )[φT ω / k * ] (8. det[ pI − Λ ] = Z m ( p ) (8. The control input can be rewritten in terms of the adjustable parameters and the various signals. Why the closed-loop transfer function can become exactly the same as that of the reference model ? To know this.14 A control system with perfect tracking The structure of this control system can be described as follows: . note that the control input in (8. The poles of the matrix Λ are chosen to be the same as the roots of polynomial Z m ( p ) . together * with the scalar gain θ 0 can move the poles of the closed-loop control system to the locations of the reference model poles.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Therefore.48) Let the ideal value of θ be θ* and the error φ(t ) = θ(t ) − θ* . define as follows θ(t ) = [k (t ) θ1 (t ) θ 2 (t ) θ 3 (t )]T ω(t ) = [r (t ) ω1 (t ) ω 2 (t ) ω 3 (t )]T Then the control law (8. Λ is an (n − 1) × (n − 1) matrix. where ω1 is an (n − 1) × 1 vector. by k m . k . With the control law (8. the control law (8. ∀t ≥ 0 . ω 2 = Λ ω 2 + hy . the high frequency gain of the plant.the reference signal r (t ) .The second part places the closed-loop poles at locations of those of reference model. and h is constant vector such that ( Λ.49) . and the ideal control parameters described above are also unknown.51) becomes u (t ) = θ T (t ) ω (t ) (8.The first part in effect replaces the plant zero by the reference model zero.52).The vector θ1 contains (n − 1) parameters which intend to cancel the zeros of plant. y 1 + W f Wu1. 8. the plant parameters are unknown. where k * . which can be described by & ω1 = Λ ω1 + hu .49) where.53) T .52) Fig. Therefore. The resulting structure of the control * system is shown in the Fig. In adaptive control problem. Instead (8.15. This is seen by noting that the transfer function from u 0 to y is Wu0 . h θ1 ω2 * Λ. θ* and θ* 2 0 represents controller parameters which lead to perfect tracking when the plant parameters are known. y = k p ( p + bm ) p + (a p1 + β1k p ) p + (a p2 + β 2 k p ) 2 (8. then θ(t ) = θ* + φ(t ) . h) is controllable.The third part of the control law (k m / k p ) r obviously replaces k p . Choice of adaptation law For the sake of simplicity. we can see the reason for assuming the plant to be minimumphase: this allows the plant zeros to be cancelled by the controller poles.The vector θ 2 contains (n − 1) parameters which.The block for generating the filter signal ω1 represent an (n − 1) th order dynamics. since the transfer function from u1 to y is Wu1. As before.Applied Nonlinear Control Nguyen Tan Tien . the control input in this system is a linear combination of: . as * u * (t ) = k *r + θ1 ω1 + θ* ω 2 + θ* y 2 0 . i. the control system with variable gains can be equivalently represented as shown in Fig. y = Wu1..49).. ym (t ) Wm ( p ) r (t ) u0 u1 e u W p ( p) k* ω1 * Λ.13 can be extended to any plant with relative degree one.e. h θ2 * θ0 since these parameters result in perfect tracking.52) can also be written as u (t ) = θ* ω + φT (t )ω .The scalar gain k * is defined to be k * = k m / k p and is intended to modulate the high frequency gain of the control system.the vector signal ω1 obtained by filtering the control input u . the output of the plant is y (t ) = B( p) * u (t ) = Wm r (t ) A( p ) (8. i.2002. __________________________________________________________________________________________ Corresponding to this control law and any reference input r (t ) .the vector signal ω 2 obtained by filtering the plant output y and the output itself. the control law is chosen to be u (t ) = k r + θ1ω1 + θ 2 ω 2 + θ 0 y (8. y = k p ( p + bm ) p + bm k p ( p + b p ) = 2 p + bp p2 + a p p + a p p + a p1 p + a p2 1 2 * . 8. with φT (t ) ω / k * regarded as an external signal. θ1 .e. At this point. y (t ) = y m (t ). θ 2 and θ 0 are controller parameters to be provided by the adaptation law. i.14.

and the polynomial D( p ) contains the parameter in the vector θ 2 . Let us define the transfer function of the feed-forward part u / u1 of the controller by λ ( p ) /(λ ( p ) + C ( p )) . since the model numerator polynomial Z m ( p ) is of degree smaller than (n − 1) . 8. Based on Lemma 8.1 and through a straightforward procedure for establishing signal boundedness. and k are chosen such that ( p + λ 0 + α 1 )( p 2 + a p1 p + a p2 ) + k p ( β 1 p + β 2 ) = ( p + λ 0 )( p 2 + a m1 p + a m2 ) and k = k m / k p . 8.13 has been replaced by a positive number λ . Noting that bm in the filter in Fig. Example 8.55) Therefore.54) ( p + λ 0 + α 1 )( p + a p1 p + a p2 ) + k p ( β 1 p + β 2 ) Since this is the familiar equation seen in Lemma 8.58) r (t ) k u0 u1 The question now is whether in this general case. Linear system with higher relative degree The design of adaptive controller for plants with relative degree larger than 1 is both similar to. we can show that the tracking error in the above adaptive control system converges to zero asymptotically. it is no longer possible to choose the poles of the filters in the controller so that det[ pI − Λ] = Z m ( p ) as in (8.57) where λ ( p ) = det[ pI − Λ ] and λ1 ( p) is a Hurwitz polynomial of degree (n − 1 − m) .2002.6 _______________________________________ Consider the second-order plant described by the transfer function y= kp p + a p1 p + a p2 2 For a general plants of relative degree larger than 1. the desired zeros of the reference model can be imposed.48). With this choice.4. Specifically. and that of the feedback part by D( p ) / λ ( p ) . 8. θ 2 and θ 0 such that the above transfer function becomes exactly the same as Wm ( p ) . such choice of parameters exists and is unique. there exists choice of values for k . Choice of control law Let us start from a simple example. the following adaptation law is chosen & θ = − sgn(k p ) γ e(t ) ω(t ) (8.2.59) u p + λ0 W p ( p) α1 β1 p + β 2 p +λ 0 The answer to this question can be obtained from the following lemma Fig. the choice of control law is quite similar but the choice of adaptation law is very different. that for plants with relative degree 1.14 is chosen. 8. Note that the order of the filters in the control law is still (n − 1) . due to the assumed positiveness of k m . 8. θ1 . h ω2 Λ.15 An equivalent control system for time-varying gains W ry = k 1+ = kp p + λ0 p + λ0 + α1 p 2 + a p p + a p 1 2 kp p + λ0 β1 p + β 2 p + λ0 + α 1 p + λ0 p 2 + a p p + a p 1 2 k k p ( p + λ0 ) 2 Since y m (t ) = Wm ( p ) r . the same control structure as given in Fig. Clearly. or equivalently R p (λ ( p ) + C ( p )) + k p Z p D( p ) = λ1Z p Rm ( p ) (8. if the controller parameters α 1 .Applied Nonlinear Control Nguyen Tan Tien . h * θ0 Fig. where the polynomial C ( p) contains the parameter in the vector θ1 . the tracking error is seen to be related to the parameter error by the simple equation e(t ) = Wm ( p ) [φT (t ) ω(t ) / k * ] (8. However.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ r (t ) u0 k φ ω k* T * u1 u * θ1 ω1 W p ( p) y Let the controller be chosen as shown in Fig. β 2 . β 1 . 8. we now choose λ ( p ) = Z m ( p ) λ1 ( p ) (8. The closed-loop transfer function from the reference signal r to the plant output y is * θ2 Λ. __________________________________________________________________________________________ where γ is positive number representing the adaptation gain and we have used the fact that the sign of k * is the same as that of k p .16. Then the closed-loop transfer function is easily found to be u and the reference model ym = km p + a m1 p + a m2 2 r ym (t ) e Wry = Wm ( p ) k k p Z p λ1 ( p ) Z m ( p ) R p ( p )[λ ( p ) + C ( p )] + k p Z p D( p ) (8. and different from.1. then the closed-loop transfer function W ry becomes identically the same as that of the reference model.16 Model-reference control system for relative degree 2 ___________________________________________________________________________________________________________ 43 Chapter 8 Adaptive Control . Instead.

Specially.Applied Nonlinear Control Nguyen Tan Tien .2 A two-link robot Consider the nonlinear dynamics of a two-link robot && & & & &  H 11 H 12   q1  − h q 2 − h q1 − h q 2   q1   g1  τ 1     &&  =  &  &  +   =   0  H 21 H 22  q 2   h q1  q 2   g 2  τ 2  (6.85) Fig. because of the present of the un-measurable joint && acceleration q . Specially. The filtered torque y can also be computed because the torque signals issued by the computer are known. _________________________________________________________________________________________ 8. 6. τ 2 I 1. 2 I 2 + m2 l c 2 .2002.84) Example 8.82) (8.7. (8.5 Least-squares with exponential forgetting ___________________________________________________________________________________________________________ Chapter 8 Adaptive Control 47 . we can use the above filtering technique. proper filter. for a linear SISO system.9) where.3 The gradient estimator 8.81) n +p 2 H11 = a3 + a 4 + a1l1 + 2a 2 l1 cos q 2 H 22 = a 4 H12 = H 21 = a 2 l1 cos q 2 + a 4 Thus we can write & && τ = Y1 (q.83) cannot be directly used for parameter estimation. m 1 q1. Then. a3 = I1 + m1lc1 . including multiple-link robots. In view of the fact that A0 ( p ) − A( p ) = (α 0 − a0 ) + (α1 − a1 ) p + K + (α n−1 − a n−1 ) p n−1 (8.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ Generally.9).7.84) can be rewritten as I2. To avoid this. let w(t ) be the impulse response of a stable. the first term on the right hand side of (8. q) a (8.τ1 l2 lc2 l1 lc1 ∫ t 0 && & w(t − r )Hq dr = w(t − r )Hq 0 − t t & ∫ dr [wH]q dr 0 t d & & = w(0)H (q)q − w(0)H[q (0)] q(0) − && & & ∫ [w(t − r )Hq .9) a4 = is linear terms of the equivalent inertia parameters a = [a1 a 2 a3 a 4 ]T . q = [q1 q 2 ]T : joint angles 2 2 2 H 11 = m1l c1 + I 1 + m 2 (l1 + l c 2 + 2l1l c 2 cos q 2 ) + I 2 where y is the filtered torque and W is the filtered version of Y1 .83) has known coefficients. m2 q 2. a 2 = m2lc 2 .2 Prediction-error-based estimation model 8. q. leading to A0 ( p ) − A( p) B( p ) u A0 ( p ) A0 ( p ) where y= A0 = α 0 + α1 p + K + α n−1 p n −1 2 H 12 = H 21 = m 2 l1c 2 cos q 2 + m 2 l c 2 + I 2 2 H 22 = m 2 l c 2 + I 2 (8.7. its dynamics can be described by A( p ) y = B ( p ) u where A( p ) = a0 + a1 p + K + a n−1 p n−1 + p n B ( p ) = b0 + b1 p + K + a n−1 p n−1 Divide both sides of (8.7.w(t − r )Hq] dr 0 This means that the equation (8. Relation (8. convolving both sides of (6. Thus the matrix W can be computed from available & measurements of q and q .4 The standard least-squares estimator τ = [τ 1 τ 2 ]T : joint inputs (torques) 8.48) can be rewritten as & y (t ) = W(q. yields & & ∫ w(t − r )τ(r )dr = ∫ w(t − r )[Hq& + Cq + G]dr 0 0 t t θ = [α 0 − a0 α1 − a1 L α n−1 − a n−1 b0 L bn−1 ] T T  y py p n−1 y u p n−1u  w= L L  A0 A0 A0 A0 A0     Note that w can be computed on-line based on the available values of y and u .9 Linear parametrization of robot dynamics_____ Using partial integration. _________________________________________________________________________________________ (8. and Then each term on the left-hand side of (6.80) by a known monic polynomial of order n.81) can be rewritten in the form y = θT w (t ) where This linear parametrization property actually applies to any mechanical system. q)a (8.80) h = m 2 l1l c 2 sin q 2 g1 = m1l c1 g cos q1 + m 2 g[l c 2 cos(q1 + q 2 ) + l1 cos q1 ] g 2 = m 2 l c 2 g cos(q1 + q 2 ) 2 Let us define a1 = m2 .

6 Bounded gain forgetting 8.6 _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ 8.Applied Nonlinear Control Nguyen Tan Tien .8 Composite Adaptation ___________________________________________________________________________________________________________ Chapter 8 Adaptive Control 48 .2002.7.7.7 Concluding remarks and implementation issues 8.