You are on page 1of 6

Cop\Ti~hl © IF.

-\C IllIh Tril'lllli,tl \I'" rld COIl~rt·",


\llIllirh. FR(;, l'IHi

CARTESIAN CONTROL OF ROBOTIC


MANIPULATORS
H. Seraji
j et Propu/sio/l La/wm/tlll'. Califomia Iw/itu/t, of Tech 11 %g:•. .,800 Oak Gm,'!'
Dri,'I'. Pa,\ adma, CA 91109 , L'SA

Abstract . A new adaptive control scheme for direct control of manipulator end-effector to achieve
trajectory tracking in Cartesian space is developed in this paper. The control structure is obtained
from lin ear multi variable theory and is com posed of simple feedforward and feedb ack controllers a nd
an auxiliary input. The direct adaptation laws are derived from moclel reference adapt ive control
theory and are not based on parameter estimation of the robot model. The utili zation of adapti ve
feed forward control and the inclusion of auxiliary input are novel features of the present scheme
and result in improved dynamic performance over existing adaptive control schemes, The adaptive
controller does not require the complex mathematical model of the robot dynamics or any knowledge
of th e robot parameters or the payload; and is computationally fast for on-line impleme ntation with
high sampling rates,

Keywords. Adaptive control; feedback ; feedforward; Lyapunov methods ; multivari able control
systems; position control; robots; tracking systems.

1. Introduction control systems which can be implemented more efficiently for


on·line controL
Although end-effector control is the ultimate goal of any
robot control system, direct control of the end-effector motion The objective of this paper is to extend the direct adap-
in Cartesian space has not attracted much attention. Conven- tive control approach presented in [20) to achieve trajectory
tionally, task description is expressed in terms of a sequence tracking of the end-effector directly in the Cartesian space.
of end-effector coordinates in the Cartesian space. This infor· The structure of the proposed control system is derived from
mation is transformed through inverse kinematics to a series linear multivariable theory; while the direct adaptation laws
of angular positions in the joint space. End-effector control
are developed using model reference adaptive control (MRAC)
is then accomplished indirectly by controlling the joint angles
theory based on the Lyapunov method.
which are related to the end-effector coordinates through for-
ward kinematics. This indirect approach to end-effector con-
trol is both computationally inefficient and unduly complicated The paper is organized as follows. In Section 2, an incre-
due to degeneracy of joint angles corresponding to a given end- mental model of end·effector motion is derived and a simple in-
effector condition. Furthermore, negligible errors in the joint cremental control scheme for trajectory tracking is developed.
angles may result in noticeable errors in the end- effector co- The results are used in Section 3 to formulate the tracking
ordinates, depending on the manipulator geometry. control problem in the MRAC framework and to derive the
adaptation laws. Section 4 discusses the results and draws
Most existing methods for robot control in the joint space some conclusions.
fall into the two major categories of ugloballinearization" and
"adaptive controL" In recent papers [1,2), Khatib has ex-
tended the global linearization method to the direct control 2. Incremental Modeling and Control of the
of end-effector motion in the Cartesian space, This method is End-Effector
based on cancelling out the nonlinearities of the robot model
by means of a nonlinear feedback controller and requires exact In this sectioll, we shall first obtain an incremental dy-
knowledge of the complex dynamic model of the robot . The namic model and t!en develop an incremental control scheme
method also needs accurate values of the robot parameters and
for small movements of the end-effector in the neighborhood
the payload . When perfect cancellations of the robot nonlin-
of some nominal operating point .
earities is not achieved due to imperfect modeling or inaccurate
parameter values, dynamic performance of the robot may be
degraded and a complicated stability analysis is necessitated Let us consider a general robotic manipulator having n
and this situation may even lead to instability of the closed- joints. Let 1:( t) be the nxl vector of torques applied to the
loop system [3 - 4). joints and S2.( t) be the nxl vector of the resulting joint angles.
The dynamic equations of motion of the robot links can be
Adaptive control offers an appealing solution to the robot represented by a set of coupled nonlinear differential equations
control problem, In adaptive robot control methods, neither of the general form [21-23)
the complex mathematical model of the robot dynamics nor
any knowledge of the robot's dynamic parameters or the pay- M(S2.)&2(t) + lY:(S2.,~) + Q(S2.) = 1:(t) (1)
load are required in generating the control action. Adaptive
control methods in general fall into two distinct categories; where the matrices M, lY: and Q are highly complex nonlinear
namely, indirect and direct . In indirect adaptive control meth· functions of S2., ~, and the payload.
ods such as [5 - 10), the paramters of the robot model are
estimated first and are then used for adjusting the controller Consider now a fixed task-related Cartesian frame of ref-
gains. In direct adaptive control methods such as [11 - 20), erence and let the nxl vector X. represent the end- effector
the control action is generated directly, without prior param- position coordinates in this frame . For a general six degree·
eter estimation. As a result, direct methods result in simpler of-freedom manipulator, X contains three position components

289
290 H. Seraji

(such as Cartesian coordinates z, y, z) and three orientation discussed separately:


components (such as Euler angles ~, .p, ~). The end-effector
position vector X is related to the joint angle vector ~ by the
forward kinematic relationship [22-23] 2. 1 Feedback Controller

X(t) = A(~) (2) The role of the ru:n feedback controller K($) is to provide
a stable closed-loop system with poles at desired locations in
the complex plane. The feedback controller also ensures that
where the ru:l vector A(~) is a nonlinear function which em-
any error between 3<.(t) and r(t) is reduced to zero asymptoti-
bodies the manipulator geometry. Equations (1)- (2) consti-
cally.
tute the nonlinear mathematical model of the end- effector
dynamics. For the purpose of stabilization and pole placement, it is
sufficient to apply feedback from the position vector 3<.(t) and
Suppose that the initial condition of the end-effector in the velocity vector ~(t); since this corresponds to full state
the Cartesian space is denoted by Xo and corresponds to the feedback for the linear model (7) . Hence, an appropriate choice
joint angle vector ~o and the joint torque vector 'L0 . Let for the feedback controller is
the operating point representing this initial condition of the
manipulator be denoted by P. Now, suppose that the joint (9)
torque vector is perturbed slightly by 1:(t), that is 'L(t) =
'L 0 +1:(t); and let the resulting perturbation in the joint angle where Kp and K. are constant ru:n position and velocity feed-
vector and the end-effector position vector be ~(t) and 3<.(t) back gain matrices respectively. The feedback control law is
respectively, i.e. ~(t) = ~o + ~(t) and X(t) = Xo + 3<.(t). In therefore given by
a recent study [20], it is shown that ~(t) is related to 1:(t) by
the linear differential equation

M(t) + B~.(t) + 6~(t) = 1:(t) (3) in the frequency-domain; or by

Equation (3) is obtained by linearizing the nonlinear equation L1 (t) = Kp~(t) + K.§.(t) (11)
(1) about the operating point P; and -4, B, and 6 are constant
ru:n matrices which depend on P. Now, consider the differen- in the time-domain, where ~(t) = r(t) - 3<.(t) and ~(t) = i(t)-
tial transformation from the joint space {~} to the Cartesian ~(t) are the nxl position and velocity tracking-error vectors
space {X}. The joint angle perturbation ~(t) is related to the respectively. Note that no differentiation is involved in imple-
end-effector perturbation 3<.(t) by [22-23] menting the feedback controller since i( t) and ~(t) [= Jo~( t)]
are directly available; and hence K($) is realizable.
(4)
2. 2 FeedCorward Controller
where J o = J(S;l0) = [1l~~;U]p is the ru:n Jacobian matrix
The function of the ru:n feedforward controller Q ($) is
evaluated at the operating point P. Assuming P is not a sin- to cause the position vector 3<.(t) track the reference trajectory
gular point of the Jacobian matrix, equation (4) can be solved vector 1:(t) . From Refs. [24-25], the feedforward controller
for ~ as Q(s) is chosen as the minimal-order inverse of the end-effector
(5) model (8); that is

In formulating the end-effector dynamics in Cartesian


space, we consider the ru:l "virtual" force vector L acting on
the end-effector as the driving input. The end- effector force
L is related to the joint torque 'L through the J acobian matrix
The feedforward control law is therefore given by
[23]
'L(t) = J'(~)L(t)
where the superscript "I" denotes transposition. Hence, for an
incremental analysis
in the frequency-domain; or by
1:(t) = J~[(t) (6)
where the ru:l vector I denotes the perturbation in the end- L 2 (t) = Cr(t) + Bi(t) + Ai(t) (14)
effector force from the nominal value L'.
in the time-domain. It must be noted that since the desired
On substituting for ~ and 1: from equations (5)-(6) into velocity i(t) and acceleration i(t) are directly available, it is
equation (3), we obtain the incremental model of the end- not necessary to perform differentiation in implementing the
effector dynamics in Cartesian space as
feedforward control law; and hence Q($) is realizable.
Ai(t) + B~(t) + C3<.(t) = L(t) (7)
Based on the foregoing results, a simple "incremental"
where A = (J')-1 Aj-1 B = (J,)-1 BJ- 1 and C = robot control system can be formed by combining the feedback
(J~)-16J;;1 are °consta.x':t ~ matric:s . Eq~ation (7) rep-
and feedforward controllers as shown in Figure 1. It must be
resents a (2n)lh order linear multivariable system with the ru:l
input vector I(t) and the ru:l output vector 3<.(t) . The transfer- noted that since we cannot physically apply Cartesian forces to
function model relating 1($) to 3<.( $) is obtained from equation the end-effector, we instead compute the joint torques needed
(7) as - to effectively cause these forces. Thus, in order to implement
3<.($) = (A$2 + B$ + C)-1[($) (8) the controllers, the control law in joint space is given by

Now, let !:(t) denote the ru:l "incremental reference tra- 1:($) = J~ [[1($) + [2($)]
jectory vector" in Cartesian space, that is, the desired value (15)
of the end-effector position vector 3<.(t) . Then, in order for the = J~ [Kp + K.$]~($) + J~ [c + B$ + A$2] r($)
position vector 3<.( t) to track the reference trajectory vector
r( t), the control system must have two independent controllers in the frequency-domain; or by
[24] as shown in Figure 1; namely the feedback controller K($)
and the feedforward controller Q ($ ). Each controller is now 1:(t) = J~ [Kp~(t) + K.§.(t) + Cr(t) + Bi(t) + Ai(t)] (16)
Cartl'si;1I1 Control oj Robotic \Llniplll ;ltors 291

in the time-domain. On applying the control law (16) to the adaptation ensures that the robot control system is tuned on-
end-effector model (7). we obtain line such that a good dynamic performance is achieved despite
coefficient variations due to changes of the robot geometric
configuration. speed of motion or the payload.

Noting that ~ = f. - i;. and £: = t -~. we obtain the (2n)th order Consider the nonlinear model of the end-effector dynam-
error differential equation ics written as [1 .2.12]

A£:(t) + (B + K.)~(t) + (C + Kpk(t) = Q (17) £:(t) = A"(X.K)X(t)+B"(X.K)K(t)+C"(X.K)X(t) (22)

Equation (17) describes the dynamic behavior of the position where A". B". and C" are nxn matrices whose elements are
tracking-error ~(t) in terms of the feedback gains Kp and K •. complex nonlinear functions of X.K. and the payload. Let us
It must be noted that by choosing the feedforward controller now consider the linear time-varying robot control law
as the minimal inverse of the robot. we have ensured that the
error differential equation (17) is homogeneous; i.e .• the right- £:(t) =£.(t) + [Kp(t)g:(t) + K.(t)t(t)]
hand side of (17) is zero. The solution of equation (17) can be (23)
expressed as + [C(t)K(t) + B(t)E.(t) + A(t)E(t)]
where g:(t) = K(t) - X(t) is the nx1 "total" position tracking-
~(t) = [:f: G;e",p('1;t)] ~(O) + [:f: H;e",p('1;t)] ~(O) (18) error vector. Equation (23) is a generalization of the "total"
,:;;::1 1::;:1 control law (21) developed in Section 2 based on the incremen-
tal analysis . As in equation (21). this control law is composed
where ~(O) and ~(O) are the initial position and velocity errors. of three components; namely the "auxiliary input" £.(t). the
G; and H; are some constant nxn matrices. and '1; is a root of
the (2n)th order error characteristic polynomial [24]
feedback term [Kp(t)g:(t) + K.(t)t(t)]. and the feedforward
.le(s) = IAs2 + (B + K.)s + (C + Kp)1 (19) term [C(t)K(t) + B(t)R(t) + A(t)E(t)] . The time-varying sig-
nal £.(t) corresponds to the operating point term £:" in equa-
tion (21) and will be synthesized by the adaptive scheme. On
It is noted that the feedforward controller Q(s) does not
applying the total control law (23) to the nonlinear end-effector
affect the error characteristic polynomial .le( s) and hence the
model (22) as shown in Figure 3. we obtain the closed-loop
dynamics of the tracking-error ~(t) is independent of Q(s) .
Since the gain matrices K p and K. can be chosen such that model
'11 •.. ·.'12n have negative real parts [24]. the tracking-error vec-
tor ~(t) in equation (18) will tend to zero asymptotically; i.e .• A"X(t) + B"K(t) + C"X(t) = £(t) + Kpg:(t) + K.t(t)
~(t) -> Q or ~(t) -> 1:(t) as t -> 00 for all ~(O) and ~(O). In + CK(t) + BR(t) + AE(t)
fact. by placing the roots '11 •.. .• '72n at desired locations. the (24)
transient behavior of the tracking-error can be shaped at the This equation can be written in terms of the total tracking-
designer's discretion. error g:( t) as
3. Direct Adaptive Control of the End-Effector
Motion A"t(t) + (B" + K.)t(t) + (C" + Kp)g:(t) = -£.(t)
(25)
+ (A" - A)E(t)+ (B" - B)R(t) + (C" - C)K(t)
In Section 2. the structure of multi variable controllers to
control the incremental motion of the end-effector was devel- It is seen that 1£( t) and £.( t) appear as forcing functions on the
oped. This control scheme works well for small movements of right-hand side of the error differential equation (25) . There-
the end- effector about the nominal operating point P. The fore. if the gains of the feedforward controller Q(s) and the
implementation of the incremental control law on the robot auxiliary input £.(t) are fixed. the solution of equation (25)
manipulator is shown in Figure 2. It is seen that the "total"
for the tracking-error g:( t) will no longer tend to zero asym-
control law is the sum of two components . The first compo-
totically. and will depend on K(t) and £.(t). As a result. it
nent is the value of the end-effector force vector at the nominal
is essential to adapt the feedforward gains and the auxiliary
operating point P; namely :£:'. The second component is the
input to cope with variations of the operating point so as to
contribution due to the incremental controllers K(s) and Q(s).
accomplish trajectory tracking. The feedback gains will also be
Thus. from Figure 2. the "total" control law is given by
adapted to ensure closed-loop stability with desired transient
£:(t) =£:0 + [(t) = :£:' + Kp [!:(t) - ~(t)] performance.
(20)
+ K. [f.(t) - i;.(t)] + C!:(t) + Bf.(t) + Af(t) Now. let us define the 2nx1 position-velocity tracking-
Now. let the "total" reference trajectory vector be K(t) = K° +
1:(t) and the "total" position vector be X(t) = XO +~(t). where error vector ;.(t) ~ (.i~g) and rewrite equation (25) in the
in practice K° = XO . Substituting these in equation (20) gives standard state-space format
the "total" control law in terms of the "total" variables as

£:(t) =E + Kp [K(t) - X(t)] + K. [R(t) - K(t)]


(21)
i(t) = (_ [AT1 ~C" + Kp] _ Wr1~B" + K.]) ;.(t)
+ CK(t) + BR(t) + A.k{t)
+ CA"r? [-£.]) + (wr 1
[~" - Cl) K(t)
where £:" = £:0 _ CKo - BRo - Ajr. It is seen that in ad-
dition to the two terms due to the feedback and feedforward + ([A"r 1 [~" - B]) R(t) + (wr 1 [~" _ A]) E(t)
controllers. a third term £:" reflecting the effect of the operat-
ing point P is present in the total control law. Note that the (26)
operating point does not satisfy the incremental model; that Equation (26) constitutes the "adjustable system" in the
MRAC framework .
is. £:0 f. [CKo + BK + AK]. i .e. E f. Q.
We shall now define the "reference model" which em-
In the direct adaptive control scheme developed in this bodies the desired performance of the system in terms of the
section. the gains of the controllers K (s) and Q (s) and the tracking-error g:(t). For the end-effector. the ,Klost desirable
operating point term £:" in the control law (21) are adapted situation is that each tracking-error E;(t) = R;(t) - X;(t) is
continuously on-line to cope with variations in coefficients of decoupled from the others and satisfies a second-order homo-
the robot model (7) due to changes in the operating point . This geneous differential equation of the standard form
292 H. Snaji

E.(t) + 2{.w.E.(t) + wl E.(t) = 0;


i = 1 , . ,n
"
(27)
6(t) = "1!I.m(t)E'(t) + "2ft [!I.m(t)R'(t)] (35)
where {. and w.
are the damping factor and the undamped
natural frequency to be specified by the designer and reflect B(t) = 1'1!l.m(t)R'(t) + 1'2ft [!I.m(t)R'(t)] (36)
the desired dynamic behavior of E.(t) . Equation (27) can be
..1(t) = Al!l.m(t)E.'(t) + A2~ [!I.m(t)E.'(t)]
written in the vector form (37)

where
where Dl = diag.(wl) and D2 = diag.(2{.w.) are constant n.xn
diagonal matrices and the subscript 'm' denotes the reference
model. Equation (28) can be put in the standard state-space
format is an nx1 vector . Let us now consider the reference model state
(29) vector ~m(t) = ( t~g) . The solution ofthe homogeneous
reference model (29) can be expressed as

where ~m(t) = ( t~g) is the 2nxl vector of desired posi- (38)


tion and velocity errors . Since the reference model (29) is sta-
where ~m(O) is the initial state of the reference model. Since
ble, there exists a constant 2nx2n symmetric positive- definite
the initial values of the reference and actual trajectories are
matrix P = (~: ~:) which satisfies the Lyapunov equation often the same, i.e. EO = KO and .11,." = XO, the initial error
is usually zero; that is ~(O) = Q. Hence, from equation (38) ,
[see Reference 20] ~(t) ;: Q for all t. Therefore, the adaptation laws (32) - (37)
simplify to
PD+D'P= -Q
E.(t) = 51!I.(t) + 521.(t) (39)
where D is the 2nx2n system matrix in equation (29) and Q is . rn '(t)
]
a constant 2nx2n symmetric positive-definite matrix . We shall Kp(t) = 0I!I.(t).E, (t) + 02did t!l.(t).E (40)
now derive the adaptation laws which ensure that the position-
velocity error vector ~(t) tends to the desired value ~m(t) for k.(t) = (31'lJt)£.'(t) + (32ft [!I.(t)£.'(t)] (41)
any reference trajectory E( t). Applying the method described
in Reference 26, we obtain
6(t) = "1!I.(t)R'(t) + "2ft l!I.(t)E'(t)] (42)

( _ [A~-1 t) = Q;;1 P~ + Q:P~ B(t) = 1'1!I.(t)E'(t) + 1'2ft [!I.(t)R'(t)] (43)

(_ [A"~-1 k p _ [A"~-1 kJ = Qil P~~' (OI;n J


{310I ..1(t) = Al!I.(t)E.'(t) + A2ft [!I.(t)E.'(t)] (44)

+ Q1.pd(
di ~~
')(02In
0 {32In
0) where the nx1 vector 'l.(t) is defined as

( _ [A~-1 6) = Q2"1 P~E' + QiPft(~R')


[A~-1 B) P~R' + Q3Pft(~R')
using the results of Reference 20. Hence
( _ = Q3"1
!I.(t) = diag.(wpi).E(t) + diag.(w •• )£(t) (45)
( _ [A~-1 A) = Q.l p~jl + Q:pft(~jr)
where w p ' = (2{.wt)g. and w •• = (wl)h. are weighting factors
(30)
of E. and E. . The constants g. and h. are chosen by the
where ~(t) = ~m(t) - ~(t) is the 2=1 adaptation error vec-
designer to reflect the relative significance of the position and
tor, {01' {31} are positive scalars, {02' {32} are zero or posi-
velocity errors E. and E•. From equations (39) - (44) , the
tive scalars, Qo, ... ,Q4 are constant 2nx2n symmetric positive-
required auxiliary input and controller gains are given by
definite matrices, and Q;, ... , Q. are constant 2nx2n symmet-
ric positive semi-definite matrices. In deriving equation (30),
we have treated the robot parameters A·(K,X), B·(K,X),
and C·(K,X) as unknown and "quasi-time-invariant" in com-
£(t) = £(0) + 52'l(t) + 51 l' !I.(t)dt (46)

+ l'
parison with the adaptation scheme. In order to cancel out
A· in the adaptation laws, the matrices in equation (30) are Kp(t) = Kp(O) + 02'l.(t).E'(t) 01 'l.(t).E'(t)dt (47)
chosen as

K.(t) = K.(O) + (32!I.(t)£' (t) + l'


(31 !I.(t)£' (t)dt (48)

C(t) = C(O) + "2'l.(t)R'(t) +"1 l' !I.(t)R'(t)dt (49)

B(t) = B(O) + 1'2'l(t)R'(t) + l'


1'1 !I.(t)E'(t)dt (50)
where {51 ,"1 ,1'1 ,Ad are positive scalars, {52, "2,1'2, A2} are
zero or POSl·t·lve sc al ars, an dA.. = (A· A.0).
0 IS a 2nx2n
A(t) = A(O) + A2!I.(t)E.'(t) + l'Al !I.(t)E.'(t)dt (51)
positive-definite matrix. Substituting from equation (31) into
equation (30) and simplifying yields the adaptation laws The control scheme is extremely simple since the con-
troller gains are evaluated from equations (46-51) by sim-
(32) ple integration using, for instance, the trapezoidal rule.
Thus the computational time required to calculate the
.
Kp(t) = olim(t).E, (t) + 02did [im(t).E'(t)] (33) adaptive control law (equation 23) is extremely short . As
a result, the scheme can be implemented for on-line robot
k.(t) = (31!I.m(t)£'(t) + (32~ [~(t)£'(t)] (34)
control with high sampling rates; resulting in improved
dynamic performance.
(:;tr(esiall COlltrol of Robotic \ialliplllatol"s

4. Conclusions 11. S. Dubowsky, and D.T. DesForges: "The application of


model-referenced adaptive control to robotic manipula-
A simple model reference adaptive robot control scheme tors," ASME J. Dyn. Systems, Measurement and Con-
using the Lyapunov method has been developed in this paper. trol, Vol. 101, pp. 193-200, 1979 .
The control scheme achieves trajectory tracking for the robot
end-effector directly in the Cartesian space. The scheme is in- 12. A . Balestrino, G. DeMaria and L. Sciavicco: "An adap-
sensitive to variations in the robot parameters and the payload, tive model following control for robotic manipulators,"
has low on-line computational burden and very low memory re- ASME J . Dyn. Systems, Measurement and Control, Vol.
quirement. Furthermore, by proper definition of the "reference 105, pp. 143-151, 1983.
model" as in Section 3, decoupled response for each degree-of-
freedom of the end-effector can be achieved . The adaptive 13. S. Nicosia and P . Tomei: "Model reference adaptive con-
control scheme presented in this paper is "direct" and gen- trol algorithms for industrial robots," Automatica, 20(5),
erates the control action without prior parameter estimation pp. 635-644, 1984 .
of the robot model. These attributes make the present scheme
suitable for implementation in on-line control ofrobotic manip- 14 . S.N . Singh: "Adaptive model following control of non-
u1ators with high sampling rates. Furthermore, the indepen- linear robotic systems," IEEE Trans. Aut . Control,
dence of adaptive control methods from the complex dynamic AC30(11), pp. 1099-1100, 1985.
model or the dynamic parameter values of the manipulator un-
der control makes such methods easily "portable" to different 15. M. Takegaki and S. Arimoto: "An adaptive trajectory
manipu1ators. control of manipulators," Int. J. Control, 34(2), pp. 219-
230, 1981.
5. Acknowledgement
16. R .P . Anex, Jr. and M. Hubbard: "Modeling and adap-
The research described in this paper was performed at the tive control of a mechanical manipulator," AS ME J. Dyn.
Jet Propu1~ion Laboratory, California Institute of Technology, Systems, Measurement and Control, Vol. 106, pp. 211-
under contract with the National Aeronautics and Space Ad- 217, 1984.
ministration .
17. M. Tomizuka and R . Horowitz: "Model reference adap-
6. References tive control of mechanical manipulators," Proc. IFAC
Symp. Adaptive Systems in Control and Signal Proc.,
1. o. Khatib: "Dynamic control of manipulators in opera- pp. 27-32, San Francisco, 1983.
tional space," Proc. 6th IFTOMM Congress on Theory
of Machines and Mechanisms, pp. 1128-1131, New Delhi, 18. K.Y . Lim and M . Eslami: "Adaptive controller designs
India, 1983 (Wiley, New Delhi). for robot manipulator systems using Lyapunov direct
method," IEEE Trans. Aut. Control, AC30(12), pp.
2. O . Khatib : "The operational space formulation in the 1229-1233, 1985.
analysis, design, and control of manipulators," Proc. 3rd
International Symposium of Robotics Researc.l, Paris, 19. M . Vukobratovic and N. Kircanski: "An approach to
France, 1985. adaptive control of robotic manipulators," Automatica,
21(6), pp. 639-647, 1985.
3. J .R. Rewit : "Decoupled control of robot movement,"
Electronics Letters (lEE), 15 (21), pp . 670-671, 1979. 20. H. Seraji: "Adaptive control of robotic manipulators,"
Jet Propu1sion Laboratory, Engineering Memorandum
4. O. Egeland: "On the robustness of the computed torque 347-182, January 1986 (Internal Document).
technique in manipulator control," Proc. IEEE Intern.
Conf. Robotics and Automation, pp. 1203-1208, San 21. A.K. Bejczy: "Robot arm dynamics and control," Jet
Francisco, 1986. Propu1sion Laboratory, Technical Memorandum 33-669,
1974 (Internal Document).
5. A.J. Koivo and T.R. Guo: "Adaptive linear controller
for robotic manipulators," IEEE Trans. Aut . Control, 22. R. Pau1: "Robot Manipulators: Mathematics, Program-
AC28(2), pp. 162-170, 1983 . ming and Control," MIT Press, Cambridge, MA, 1981.

6. C.S.G. Lee and M.J. Chung: "An adaptive control strat- 23. J .J . Craig: "Robotics: Mechanics and Control,"
egy for mechanical manipulators," IEEE Trans . Aut . Addison- Wesley Publishing Company, Reading, MA,
Control, AC29(9), pp. 837-840, 1984. 1986.

7. T .H. Guo and A.J. Koivo : "On a linearized model and 24. H . Seraji: "Line"! multivariable control of robot manip-
adaptive controller implementation for manipulator mo- ulators," Proc. IEEE Intern. Conf. Robotics and Au-
tion," Journal of Robotic Systems, 1(2), pp . 141-156, tomation, pp. 565-571, San Francisco, 1986.
1984.
25. H. Seraji: "In verses for multi-link robots with applica-
8. A.J. Koivo: "Self-tuning manipulator control in Carte- tion to trajectory tracking," Proc. 19th Asilomar Conf.
sian base coordinate system," AS ME Jour. Dyn. Sys- Circuits, Systems and Computers, pp. 139-143, Pacific
tems, Measurement and Control, Vol. 107, pp. 316-323, Grove, 1985.
1985.
26. H. Seraji: "A simple method for model reference adaptive
9. C.S.G. Lee and B.H. Lee: "Resolved motion adaptive control," Submitted for publication, 1986.
control of mechanical manipulators," ASME J. Dyn. Sys-
tems, Measurement and Control, Vol. 106, pp . 134-142,
1984.

10 . C.S.G. Lee, M.J. Chung, and B.H. Lee: "An approach


of adaptive control for robot manipulators," Journal of
Robotic Systems, 1(1), pp. 27-57, 1984.
294 H . Se raji

f2
Q(s) • C + Bs + As2

r-- ,..---
fl + f
r + -o-!. K(s) • K + K s
P ~()--!-. J' r!-.. MANIPULATOR
~ Jo x
,- v + 0
'---
DYNAMICS
-

Figure 1. Incremental Robot Control System

C + Bs + As 2

K +K s
P V

L__ _ _ _ _ _ _ _ _ ---1

Figure 2. Implementation of Robot Control System

R +

""0
.....
<
""..........z
~
:>
X
""
0
..... X
u
.....
< R
""
.....
z
<
V>
~
""
<
u

R +

Figure 3. Adaptive Robot Control Scheme

You might also like