1

Global Positioning of Robot Manipulators with Mixed Revolute and Prismatic Joints
Josip Kasac, Branko Novakovic, Dubravko Majetic and Danko Brezak Faculty of Mechanical Engineering and Naval Architecture University of Zagreb, I. Lucica 5, HR-10000 Zagreb, Croatia e-mail: {josip.kasac, branko.novakovic, dubravko.majetic, danko.brezak}@fsb.hr Phone: +385-1-6168-{375, 354, 348, 357} Fax: +385-1-6168-351 Corresponding author: Josip Kasac

Abstract— The existing controllers for robot manipulators with uncertain gravitational force can globally stabilize only robot manipulators with revolute joints. The main obstacles to the global stabilization of robot manipulators with mixed revolute and prismatic joints are unboundedness of the inertia matrix and the Jacobian of the gravity vector. In this paper a class of globally stable controllers for robot manipulators with mixed revolute and prismatic joints is proposed. The global asymptotic stabilization is achieved by adding a nonlinear proportional and derivative term to the linear PID controller. By using Lyapunov’s direct method, the explicit conditions on the controller parameters to ensure global asymptotic stability are obtained. Index Terms— Position control, manipulators, robot dynamics, stability.

I. I NTRODUCTION It is well known that a PD plus gravity compensation controller can globally asymptotically stabilize a rigid-joints manipulator with both revolute and prismatic joints [1]. This approach has drawbacks since the gravitational torque vector, which depends on some usually uncertain parameters, is assumed to be known accurately. To overcome the parametric uncertainties on the gravitational torque vector, an adaptive version of PD controller has been introduced in [2], guaranteeing global asymptotic stability. The main drawback of this approach is that the gravity regressor matrix has to be known. On the other hand, most industrial robots are controlled by linear PID controllers which do not require any component of robot dynamics into its control law. A simple linear and decoupled PID feedback controller with appropriate control gains achieves the desired position without any steady-state error. This is the main reason why PID controllers are still used in industrial robots. The local asymptotic stability of the linear PID controller in a closed loop with robot manipulator is proved in [3],[4] and semiglobal asymptotic stability is shown in [5], [6]. By looking at the proof, it can be seen that the cubic term in the derivative of the Lyapunov’s function hampers the global asymptotic stability. This is the reason to believe that linear PID control is inadequate to cope with highly nonlinear systems like robot manipulators, since the design of the linear PID control law is based solely on local arguments. The first nonlinear PID controller which ensures global asymptotic stability (GAS) is proposed in [7]. In this work,
This work was supported by the National Scientific Foundation of Republic of Croatia under Grant No. 0120025, ”Application of Artificial Intelligence in Robotics and Manufacturing Systems”.

which was inspired by the results of [2], it is proven that global convergence is still preserved if the regressor matrix is replaced by the constant matrix. Since the regressor matrix is constant, the control law can be interpreted as a nonlinear PID controller which achieves GAS by normalized nonlinearity in the integrator term of the control law. The second approach to achieving GAS is the scheme of Arimoto [8] that uses a saturation function in the integrator to render the system globally asymptotically stable, just as the normalization did in [7]. A unified approach to both above mentioned controllers, which have a linear derivative term, linear or saturated proportional term, and a class of nonlinear integral action is given in [9]. An alternative approach to global asymptotic stabilization of robot manipulator is ”delayed PID” (PId D) [5]. PId D can be understood as a simple PD controller to which an integral action is added after some transient of time. The idea of this approach consists of ”patching” a global and a local controller. The first drives the solutions to an arbitrarily small domain, while the second yields local asymptotic stability. None of the proposed controllers takes into account the effects of actuators saturation. Recently, several saturated PID controllers for robot manipulators with uncertain gravitational force have been reported: a semiglobal saturated linear PID controller [10] and two global saturated nonlinear PID controllers [11], [12]. Also, progress is made in the adaptive set point control of robot manipulators. In [13] an adaptive set point controller with uncertain gravity regressor is proposed and global asymptotic convergence is proved. In [14] a semiglobal adaptive controller with amplitude-limited torque and uncertainty in the kinematic and dynamic models is developed. All of the mentioned nonlinear PID controllers can globally or semiglobally stabilize robot manipulators with revolute joints only. In this paper an approach to GAS of robot manipulators with mixed revolute and prismatic joints is presented. In this approach GAS is achieved by adding a nonlinear proportional and derivative term to the linear PID controller. Explicit conditions on controller parameters which guarantee GAS are given. Throughout the paper we use the notation: x for the Euclidean norm of the vector x ∈ Rn , λM {A} and λm {A} for the maximal and minimal eigenvalues, respectively, of the symmetric positive definite matrix A, and I for the identity matrix of the appropriate dimension. This paper is organized as follows. Robot dynamics and its

g. q) is skew-symmetric. ¯ ¯ kc = c1 + d1 qd . There exist such positive constants kg1 and kg2 that the Jacobian of the gravity vector satisfies ∂g(q) ≤ kg1 + kg2 q . a2 > 0. First. (16) ¯ where kg is defined by (9) and m = a2 + c2 qd + d2 qd 2 . ROBOT DYNAMICS The model of n-link rigid-body robotic manipulator. However. kg2 = 0. c2 . q)q + g(q) = u. ˙ Property 4. where c1 . Property 3. (9) −1 then the equilibrium [˜T q T (ν + KI g(qd ))T ]T = 0 is q ˙ globally asymptotically stable. (1) kD (13) (4) for all z. ˙ i. ˙ ˜ (11) where KP . q ∈ Rn . Then. [6]. and ˜ ¯ kg = kg1 + kg2 qd . The Coriolis and centrifugal terms C(q. k1 > max λM {KI } ¯ m1 + kc d1 + d2 ¯ m ¯ . . q is ˙ the n × 1 vector of joint velocities. where a1 . So. d2 . [15]. with the uniform bounds within the robot workspace the entire stability analysis becomes local.e. ∀ q ∈ Rn . In that case we get well known properties of the robot manipulators with revolute joints only. q)q ˙ ˙ satisfy C(q. Finally. ¯ m1 = c2 + 2d2 qd . a global result is always more useful and exact characterization of upper bounds in (4)-(6) for all q ∈ Rn is crucial in establishing global stability. It is often very difficult to explicitly characterize the domain of attraction that could be much smaller then the robot workspace. . q ∈ Rn . then d1 .. c1 = kc and kg1 = kg (see e. (See the Appendix A) For any constant vector of desired joint positions qd ∈ Rn . ˜ q I+ ˜ (2) kD (12) q I. ∂q (2) The assumptions (4)-(6) are valid for all practically used robot manipulators. Main Result Proposition 1: Consider the robot dynamics (1) in closed loop with nonlinear PID controller u = −KP q − KD q − KI ν − ΨP (˜)˜ − ΨD (˜)q. c2 .2 main properties are presented in Section II. the prismatic joints of robot manipulators are always limited by physical limits and the uniform bounds exist in (4)-(6) within the robot workspace. B. The matrix M (q)−2C(q. and it is necessary to know the domain of attraction. obtained as the gradient of the robot potential energy U (q) g(q) = ∂U (q) . and can be divided in four parts. II. u is the n × 1 vector of applied joint torques and forces. If the following conditions are satisfied kP > kg2 . and the inertia matrix and the Jacobian of the gravity vector are uniformly bounded. error equations for the closed-loop system (1). ¯ k1 = λm {KP } − kg > 0. q ˜ ˜ (8) are globally positive definite with respect to q ∈ Rn . Second. ΨP (˜) and ΨD (˜) are n × n positive q q definite diagonal matrix functions which can be written in the following form ΨP (˜) = kP q ΨD (˜) = q (1) kP . the concluding remarks are emphasized in Section V. Proof of Main Result The stability analysis is based on Lyapunov’s direct method. From a practical point of view. ∂q (6) where and are positive constants. ˜ 2 ≤ z T M (q)z ≤ (a2 + c2 q + d2 q 2 ) z 2 . A C LASS OF N ONLINEAR PID C ONTROLLERS A. [20]). If the robot has no prismatic joints. where a1 = λm {M }. q)q ≤ (c1 + d1 q ) q 2 . C(q. KD and KI are n × n constant positive definite symmetric matrices.g. [16]. [19]. in the absence of friction and disturbances. the Lyapunov function (LF) candidate is proposed. the functions 1¯ 1 G(˜) = kg q 2 + kg2 q 3 + U (q) − U (qd ) − q T g(qd ). ˙ z T (M (q) − 2C(q. The inertia matrix M (q) is a positive definite symmetric matrix which satisfies a1 z 2 q I. q))z = 0. and (11) are determined. is represented by M (q)¨ + C(q. (10). q ˙ ˙ (1) where q is the n × 1 vector of robot joint coordinates. Finally. III. where a class of PID controllers with nonlinear proportional and derivative gains is introduced and conditions for global asymptotic stability are established. ˙ Property 1. as pointed in [16]. M (q) is n × n inertia matrix. ˙ (3) Property 2. (17) (18) (19) The property (6) follows from the fact that the potential energy depends linearly on translational coordinates. (7) q ˜ ˜ ˜ 2 3 2 3 T ¯ ˜ D(˜) = kg q + kg2 q + q (g(q) − g(qd )). (1) kD (2) kD (1) We assume that the matrix C(q. [18]. [17]). the LaSalle invariance principle is invoked to guarantee the asymptotic stability. Property 5. q)q is the n×1 vector of centrifugal and Coriolis torques ˙ ˙ and g(q) is the n×1 vector of gravitational torques and forces. ∀z ∈ Rn . ˙ ˙ ˙ (5) for all q. where ˜ q = q − qd denotes the joint position error vector. while rotational coordinates appear only in trigonometric form with period 2π. The simulation results are presented in Section IV. (10) ˜ ˙ q q q ˙ ν = q. d2 ≥ 0. The following properties of the robot dynamics with mixed revolute and prismatic joints are important for stability analysis (see e. a global stability criterion on system parameters is established. q) is defined using the ˙ Christoffel symbols. d1 ≥ 0. The main results are presented in Section III. a2 = λM {M }. (1) (2) λm {KD } kD kD (1) (14) (15) .

q) = q ˙ q ˙ q q ˙ W1 (˜. the following decompositions are made: V (˜. q (27) In this way. where q ˙ q 1 1 ˙ ˙ q ˙ q ˜ V1 (˜. q = 0. Comparing (34) with (31). First.3 1) Error Equations: The stationary state of the system (1). ˜ ˜ (32) α Applying property (7) we get V2 ≥ 1 1 k1 − λM {KI } 2 α + G(˜) ≥ 0. and ν ∗ satisfies g(qd ) = ˜ ˙ −KI ν ∗ . z) is the Lyapunov function candidate (see the q ˙ Appendix B). The following step is the condition which ensures that the time derivative of LF is a negative semi-definite function. First. q ˙ (23) dt where V (˜. respectively. Note that. ν = ν ∗ .e. ¯ ¯ The function V1 is positive definite if the following conditions are satisfied λm {KD } > α. q ˙ ˙ (26) T W2 (˜) = q (αKP − KI )˜ + q ˜ q + αkP (1) where m and m1 are defined by (17) and (18). the problem of determining conditions for positive definiteness of function V (˜. q q ˜ 2 + 1 (1) k − kg2 3 P q ˜ 3 + (33) which is positive definite if the conditions α> λM {KI } . including condition (14). where k1 is defined by (15). q ˙ ˙ u = −KP q − KD q − KI z − ΨP (˜)˜ − ΨD (˜)q. ˜ ˜ (24) 3 4 1 1 V2 (˜. then the system (1). ¯ > λM {KI }d2 . If a new variable z = ν − ν ∗ is introduced. Applying properties (4) and (5) we get W1 ≥ (λm {KD } + kD q + kD q 2 ) q ˜ ˜ ˙ − α(a2 + c2 q + d2 q 2 ) q 2 − ˙ − α(c1 + d1 q ) q q 2 ≥ 0. we consider function V2 . the following conditions for positive definiteness are obtained k1 λm {KD } > λM {KI }m. q. we consider function V1 . q)q + g(q) − g(qd ) = u. q))q. 2) Lyapunov Function Candidate: We will now construct a Lyapunov function for the system (20)-(22). ˙ q 2 − (29) Using the triangle inequality q ≤ q + qd . z) = αz T KI z + q T KI z + q T KP q + q ˜ ˜ ˜ 2 2 1 (1) + kP q 3 + U (q) − U (qd ) − q T g(qd ). (11) becomes M (q)¨ + C(q. q ˜ ˜ ˜ (28) 2 3 4 and (14) are satisfied. d2 (2) (31) Further. z) and W (˜. The q ˙ q second advantage of the above mentioned decomposition of functions V and W is the elimination of unspecified constant α from the final stability conditions. which can be rearranged in the following form V2 = T √ 1 1 1 √ αz + √ q ˜ KI αz + √ q + ˜ 2 α α 1 (1) 1 + kP q 3 + q T KP q + U (q) − U (qd ) − q T g(qd ) − ˜ ˜ ˜ ˜ 3 2 1 − q T KI q . in the above-stated conditions. which can be rearranged in the following form 1 1 T ˙ q ˙ q ˜ q V1 = (q + α˜) M (q) (q + α˜) − α2 q T M (q)˜ + 2 2 1 1 (1) 1 (2) + α˜T KD q + αkD q 3 + αkD q 4 . ¯ (1) k1 kD (2) k1 kD (35) (36) (37) > λM {KI }m1 . is transformed into two simpler problems of determination of conditions for positive definiteness of functions V1 (˜. which contains three q ˙ variables. z). z) = V1 (˜. W ≥ 0. the unspecified positive constant α is eliminated. and ˙ q the inner product between (20) and y is made. q. we consider function W1 . q). (10). q) = q T M (q)q + α˜T M (q)q + α˜T KD q + q ˙ 2 2 1 (1) 1 (2) + αkD q 3 + αkD q 4 . 3) Stability criterion determination: The following step is determination of conditions for positive definiteness of the function V and positive semi-definiteness of the function W . z) q ˙ = −W (˜. For easier determination of conditions for positive definiteness of function V and W . (10). q) + W2 (˜). resulting in a nonlinear differential form which can be separated in the following way dV (˜. m ¯ kD > α. i. which contain only two variables. q) + V2 (˜. m1 ¯ (1) kD > α. z). ˙ ˜ (20) (21) (22) and using property (4) we get V1 ≥ 1 (1) (2) α(λm {KD } + kD q + kD q 2 ) q ˜ ˜ ˜ 2 1 ˜ − α2 (a2 + c2 q + d2 q 2 ) q 2 + 2 2 + a1 q + α˜ ≥ 0. ˜ ˙ q q q ˙ z = q. First. q. q. and rearrang˜ ing the previous expression we get V1 ≥ 1 1 (1) α(λm {KD } − αm) q 2 + α(kD − αm1 ) q 3 + ¯ ˜ ¯ ˜ 2 2 1 (2) + α(kD − αd2 ) q 4 + a1 q + α˜ 2 . q) = −αq T M (q)q + q T (KD + ΨD (˜))q + q ˙ ˙ ˙ ˙ q ˙ ˙ + α˜T (M (q) − C(q. ˜ ˙ (1) (2) 2 − (38) . q) and V2 (˜. k1 (34) q ˜ 3 + α˜T (g(q) − g(qd )). (25) ˜ ˜ 3 and W1 (˜. ˜ ˙ q (30) 2 whose origin [˜T q T z T ]T = 0 ∈ R3n is the unique q ˙ equilibrium. (11) is q = 0. an output variable y = q + α˜ with some α > 0 is introduced.

m2 g sin q1 q (τ )dτ. λm {KD } k (1) D . From the stability conditions it follows that the choice of the controller gains depends on the norm of the desired position qd . shows the position errors q1 and q2 obtained from ˜ ˜ simulation. The function W1 is positive definite if the following conditions are satisfied λm {KD } > α.8 m. c2 . (1) (1) (2) KI = diag{300. ¯ (1) ¯ k1 kD > λM {KI }(m1 + kc ). 200}. 2m2 g}. The control law (10)-(13) can also be applied to robot manipulators with all revolute joints.1 kg m2 . ¯ (2) k1 kD The robot manipulator with revolute and prismatic joints. The matrices M (q). q) = ˙ g(q) = m2 q 2 q 2 ˙ −m2 q1 q2 ˙ m2 q 1 q 2 ˙ . by invoking the LaSalle invariance principle. in the above-stated conditions. (d1 + d2 ) > α. 20}. (48) and (49) in expressions (4). q 2 (1) kD (2) kD lc1 x1 m2 g q1 +α (1) kP − kg2 q ˜ 3 + (41) Fig. d2 vanish and the following simplified control law u = −KP q − KD q + kD q q − KI ˜ ˙ ˜ ˙ (1) t 0 . The controller gains are chosen in accordance with stability conditions (14)-(16) as KP = diag{200. kD = 20. Finally. are the final stability conditions which guaranty global stability. The larger values of the robot position q increase the upper bounds of the mentioned matrices. Using property (8) we get W2 ≥ (αk1 − λM {KI }) q ˜ + αD(˜) ≥ 0. The conditions (42)-(44) can be represented in a more compact form by the expression (16). C(q. The final desired positions are q1d = π/2 rad and q2d = 2 m. Replacing qd from expressions (9). In other words. KD = diag{20. So. the unspecified positive constant α is eliminated. m2 (47) (48) (49) C(q. the control law (45) provides an alternative approach to the set-point control of robot manipulators with all revolute joints without using saturation functions. kD = 10. larger gains should be chosen if the desired position qd is larger. 1. Notice that the conditions (35)-(37) are trivially implied by the conditions (42)-(44). if the following condition is satisfied λm {KP } − kg > max λM {KI } λM {M } kc . c2 = 0. It is obvious from the figure that. m2 = 1 kg. 300}. [16]. . parameters kg2 . d2 = m2 . (40) m ¯ (m1 + kc ) ¯ Further. d1 . ˙ ¯ where kc is defined by (19). including (14)-(15). 2. kg1 = max{2m1 lc1 g. implying also larger values of the controller gains in the stability conditions.4 Using triangle inequality q ≤ q + qd . Also. 0 m1 lc1 g cos q1 + m2 gq2 cos q1 . q1 is the revolute joint variable and q2 is the prismatic joint variable. (17)-(19) with the norm of the maximal desired position of robot manipulators qd M . Remark 1. S IMULATION E XAMPLE We consider a 2-DOF manipulator with revolute and prismatic joints [22]. M (q) = m1 g y0 (42) (43) (44) So. lc1 = 0. as shown in Fig. and rearranging ˜ the previous expression we get W1 ≥ [λm {KD } − αm] q 2 + ¯ ˙ (1) ¯ + [kD − α(m1 + kc )] q q ¯ ˜ ˙ + (2) [kD 2 2 x0 q2 y1 + (39) − α(d1 + d2 )] q ˜ 2 q . we conclude asymptotic stability. the position errors tend asymptotically to zero. ¯ > α. after a transient due to error in initial condition. d1 = 2m2 . Fig. the conditions (42)-(44). and ˙ ˙ q2 (0) = 1 m. Remark 2. 2 a2 = max{m1 lc1 + I1 + I2 . I1 + I2 = 0. In that case. Inserting (47). This property of the stability conditions is consequence of the fact that the inertia matrix and the Jacobian of the gravity vector are not uniformly bounded. which is positive definite if the conditions (34) and (14) are satisfied. (46) where q = [q1 q2 ]T . ˜ 0 (45) provides global asymptotic stability [21]. kg2 = 2m2 g. we get the stability conditions which are valid for any qd in the robot workspace. (5) and (6) and using definitions and properties of matrix norms. q) and vector g(q) in (1) are given by ˙ 2 2 m1 lc1 + m2 q2 + I1 + I2 0 > λM {KI }(d1 + d2 ). we consider function W2 . IV. including condition (14)-(15). m2 }. kP = 20. Comparing (34) with (40) the following conditions for positive definiteness are obtained k1 λm {KD } > λM {KI }m. Initial conditions are q1 (0) = q1 (0) = q2 (0) = 0. The numerical values of the model parameters are: m1 = 1 kg. ... we get √ c1 = 0. 1.

5 −2 0 Fig. q)q + q T (g(q) − g(qd )) + ˙ q ˙ ˙ ˙ ˙ T T +q KP q + q KD q + q T KI z + ˙ ˜ ˙ ˙ ˙ +kP (1) −1. following ideas reported in [11]. (52) .5 −1 q − q (rad) 1 1d q − q (m) 2 2d for all z. q ˜ q ˙ q (53) Some terms in the differential form (53) can be represented in the following way q T M (q)¨ = ˜ q q T M (q)¨ = ˙ q q T Kj q = ˙ ˜ q T KI z = ˙ q T KI z = ˜ q k qT q = ˜ ˜ ˙ d T q M (q)q − q T M (q)q − q T M (q)q.5 0 −0. 2. function G(˜) is a globally strictly convex funcq tion vanishing at the unique global minimum q = 0. Some preliminary results guarantee only semiglobal asymptotic stability. ˜ ˜ (56) dt 2 d T q KI z − q T KI q . (60) ˜ dt where we used (2). The second problem is the extension of the proposed controller with actuator constraints and velocity estimation through the filtered position measurement. q)q + α˜ (g(q) − g(qd )) + q q q ˙ ˙ q +α˜T KP q + α˜T KD q + α˜T KI z = 0. q A PPENDIX B This Appendix presents the construction of the Lyapunov function V = V1 + V2 defined by (24) and (25).5 0. thus q ˜ G(˜) has a critical point at q = 0. C ONCLUSIONS In this paper a class of globally stable controllers for robot manipulators with mixed revolute and prismatic joints has been presented. The stability criterion in terms of Lyapunov’s direct method is proposed to guarantee the global asymptotic stability. The gradient F(˜) vanishes at q = 0. The stability criterion provides explicit conditions on the controller gains in terms of a few parameters extracted from the robot dynamics. i. positive definiteness of the Hessian H(˜) implies that the gradient F(˜) is a strictly q q increasing vector function with respect to q . q ˜ ∂[˜ + qd ] q The Hessian is a positive definite matrix because of ¯ z T H(˜)z = (kg + kg2 q ) z q ˜ ≥ kg1 + kg2 q − 2 (51) By inserting expressions (54)-(60) into (53) and separating terms in the form of the time derivatives (first terms on the right sides of equations (54)-(60)) on the left side of the equality and the rest of the terms on the right side. using definition ˙ ˜ (2) we get the following identity q T (g(q) − g(qd )) = ˙ d U (q) − U (qd ) − q T g(qd ) . (58) dt 2 1 d q k+2 . Of course. ˙ ˙ ˙ ˙ ˙ (55) dt 2 2 d 1 T q Kj q . Further. j = P. ˜ (59) dt k + 2 where we used z = q in (57) and (58). + kg2 ∂g(q) ∂q (z T q )2 ˜ ∂g(q) + zT z q ˜ ∂q z 2 ≥ 0. k = 1. (54) ˜ ˙ ˙ ˙ ˜ ˙ ˙ dt d 1 T 1 q M (q)q − q T M (q)q. ˜ ˜ ˜ (57) dt d 1 T z KI z . and property (6). D. ˜ Therefore.e. ACKNOWLEDGMENT The authors would like to thank the anonymous reviewers for their helpful comments and suggestions. q T F(˜) = ˜ ˜ q D(˜) ≥ 0. Further. (50) ˜ ˜ ˜ q ∂q ˜ q q T q + kD q q ˜ ˙ ˜ ˜ ˙ 3 (1) αkD T T (1) 2 + kD q ˜ (2) αkD T (2) 2 q ˙ 2 T 2 + (1) +αkP T q + ˜ q q q+ ˜ ˜ ˙ q q q+ ˜ ˜ ˙ +α˜ M (q)¨ + α˜ C(q. A PPENDIX A This Appendix presents a proof of Property 5 following ideas reported in [2] and [9]. The first one is the evaluation of the robustness of the proposed controller in the face of measurement noise and joint flexibility. 2 Time (sec) 3 4 V. where we used triangle inequality q ≤ q + qd . we get the decomposition (23). 2. [5]. Notice first that G(0) = 0. 1 The position errors. The Hessian matrix of G(˜) q ˜ q with respect to q is given by ˜ q qT ˜˜ ∂g(˜ + qd ) q ¯ H(˜) = kg I + kg2 q I + kg2 q ˜ + . Taking inner product between (20) and output variable y = q + α˜ we get ˙ q the following nonlinear differential form q T M (q)¨ + q T C(q. some open problems remain to be solved and are currently under research. q ∈ Rn . The gradient of G(˜) with respect q to q is given by ˜ F(˜) = q ∂G(˜) ¯ q = kg q + kg2 q q + g(˜ + qd ) − g(qd ). This ˜ implies that G(˜) is a globally positive definite function which q holds for any constant qd .

pp. [16] R. no.” IEEE Trans. “A tuning procedure for stable PID control of robot manipulators. Hernandez. [21] J. Autom. Takegaki and S. P.. pp. Sira-Ramirez. pp. “On the boundedness of the Hessian of the potential energy of robot manipulators. 141–148. 1998. “Saturated nonlinear PID global regulator for robot manipulators: Passivity based analysis. Autom. MA. London: Springer-Verlag. “Comments on: Adaptive PD control of robot manipulators. Ghorbel. Brezak. 1997.” in Proc. [3] S. pp.” Stability and Control: Theory and Applications. Autom. pp.” Int. Control.. R. 1995. Soc. “Asymptotic stability and robustness of PID local feedback for position control of robot manipulators. Tomei. “Adaptive SP-D control of a robotic manipulator in the presence of modeling error in a gravity regressor matrix: theory and experiment. 117–119.” J. Kelly. J. 22–28. and H. “Global asymptotic stability of robot manipulators with linear PID and PI2 D control. Dixon.” ASME J. and I. Santibanez. [18] R. 1288–1298. [2] P. [8] S. V. Syst. A. Robot. [9] R.. Lefeber. [7] R. pp. no. Loria. Amer. 39. 2000. 3. Kelly. [15] S. vol.” Dynamics and Control. C. no. ASME International Mechanical Engineering Congress and Exposition. 2. no. Naniwa. September 1990.” in Proc. 203–222. Dyn. no. Czech Republic. Kelly. 934–938. 2002. vol.” in Proc. Novakovic. Autom. Nijmeijer. vol. [14] W. Alvarez. Gunawardana. 61–72. Robotic Syst. no. Oxford University Press. Control of Robot Manipulators in Joint Space. and H. 18. Stoten. 1999. 16th IFAC World Congress. Conf. “Adaptive PD control for robot manipulators. “Performance tuning for a new class of globally stable controllers for robot manipulators. 1005–1012.” IEEE Trans. 2005. vol. 3. 1. Kelly. Passivitybased control of Euler-Lagrange Systems: Mechanical. Ghorbel and R. 565–570. pp. and H. Santibanez. Arimoto. pp. pp. 14. Ghorbel and R. 1994. 382–386. D. “A new feedback method for dynamic control of manipulators. Meas. [12] M. 119. [13] H. V. Loria. [4] R. 4-8 July 2005. 43. Control Conf. Int. Dyn. no. vol. vol. Majetic. on Automation. and D.” in Proc. pp.” Syst. A. 38. Contr.. “Robust stabilization of robotic manipulators by PID controllers. Arimoto. “Semiglobal stability of saturated linear PID control for robot manipulators. “A uniform bound for the Jacobian of the gravitational force vector for a class of robot manipulators. 1993.” ASME J. Control Lett. 1997. vol. Prague. [20] S.” Automatica. vol. Gorez. 1999. pp. Freidovich. Engg. 3839–3844. 1981. 613–625. Nicklasson. Arimoto. pp. vol. Singapore. no. Cervantes. Syst. Suzuki. and A. 103. San Francisco. 7. Robot. [11] R. “A class of quasi-natural potentials and hyper-stable PID servo-loops for nonlinear robotic systems. pp. E. Robot.” Trans. Luis. Loria. . Kasac. Gunawardana and F. no. Control. vol. Yazarel. Prague.. T. [17] F. Control Theory of Nonlinear Mechanical Systems: A Passivity-Based and Circuit-Theoretic Apprach. 4-8 July 2005. pp.. 9. 989– 995.. London: Springer-Verlag. vol. 13. “Globally stable PID-like control of mechanical systems.6 R EFERENCES [1] M. Robotics.” in Proc. 110–114. Liaw. P. 9.. 1999. Pervozvanski and L. 1991. “On the uniform boundedness of the derivative of the gravitational force of serial link robot manipulators. 9. 1998. 4.” IEEE Trans. Electrical and Electromechanical Applications. 3. pp. 7. pp. vol. B. [22] F. C. 11. B. 16th IFAC World Congress. Arimoto. J. vol. [19] R. Instrument Contr. Boston.” IEEE Trans. 2003. 72. 30. [10] J. 373–379. “Adaptive regulation of amplitude limited robot manipulators with uncertain kinematics and dynamics. 16. [6] A. Meas. and Computer Vision. 1999. California. Cheah. Gunawardana.” Robotica. Hodgson and D. [5] A. and H. P. Kelly. “Global positioning of robot manipulators via PD control plus a class of nonlinear integral actions. 4-6 June 2004. pp. 119–125. Contr. 138–149. Ortega. C. vol. November 12-17 1995. and M. “Robustness of the minimal control synthesis algorithm to non-linear plant with regard to the position control of manipulators. V. Czech Republic..

Sign up to vote on this title
UsefulNot useful