Adaptive control of robot manipulators including motor dynamics

An adaptive control scheme for robot manipulators including motor dynamics is proposed in this paper. The proposed scheme avoids the assumption that the values of motor parameters are known which is required in reference (13). An exponential control law is first developed under the assumption of no uncertainty. This forms a controller structure for the adaptive control. Using this control structure, a full-order adaptive control law is proposed to overcome parameter uncertainty for both robot link and motor. The stability analysis is in the Lyapunov stability sense. The method is further extended to the task space. Extensive simulations are performed to compare the direrent control schemes. Key words: adaptive control, robot manipulators, motor dynamics, exponential control law, stability analysis, manipulator dynamic equations L NOTATION A belongs to B positive constants vector of centrifugal and Coriolis forces unknown parameter vector inertial matrix ith element of vector g(q) ijth element of matrix D(q) coefficients of centripetal and Coriolis forces back electromotive force (e.m.f.) of joint j vector gravity forces ith motor or ith joint (subscript) armature current ofjth joint reflected actuator inertia matrix of the joint inertia of link i with respect to its own coordinate system n x n identity matrix jth motor or jth joint (subscript) Jacobian matrix positive definite linear feedback matrix motor back e.m.f. constant matrix motor torque constant matrix positive definite gain matrices positive definite adaptation gain matrices armature circuit inductance matrix = K;~N-~L mass of link i = Wq) + I n Y- Y d Thc MS was received on 17 August 1994 und was accepted for publication o