Adaptive generalized model-based control of robot manipulators

摘要

An adaptive model-based control approach to the trajectory tracking of robot manipulators is proposed, which incorporates both the full-order actuator dynamics and the second-order manipulator dynamics in controller design. Flexibility of design and trade-off between tracking precision and control law complexity are provided: depending on the complexity of the manipulator and of the application, the manipulator dynamics can be ignored as a disturbance, or compensated either offline or online. The proposed controllers include, as special cases, the independent joint PD control, feedforward control and computed-torque control laws derived from the manipulator model-based control approach. Simulation results are presented for the first three joints of the PUMA 560 arm, illustrating the effectiveness of the proposed controllers.