Discrete computed-torque sliding-mode manipulator control: formulation and experimental validation

A sliding-mode trajectory controller for robotic manipulators is designed using manipulator dynamics and experimentally tested using a multiprocessor architecture and a PUMA 550 robot arm. Decentralized control filters are built around actuator models while global control filters are designed around manipulator dynamics. The controller is made up of a nominal component designed using a computed torque approach and a correction component based on a variable structure suction control approach. The second control component is designed using bounds on the difference between the used and actual values of model parameters. Since the continuous manipulator system is digitally controlled along a trajectory, a discretized equivalent model of the manipulator is used to develop the controller. To facilitate the implementation of the developed controller in real time, both the nominal and correction components of the nonlinear dynamics-based feedback signal are computed using the Newton-Euler equations of motion