GA BASED COMPENSATING LEARNING CONTROL FOR FREE-FLOATING SPACE ROBOT SYSTEM WITH DUAL-ARMS IN JOINTS SPACE

In this paper, the kinematics and dynamics of thefree-floating space robot system with dual-arms are analyzed. Theclosed-loop dynamic error equation considering uncertainties for such systemis derived by using the PD type computed torque method. A new learningcontrol method using real coded genetic algorithm (GA) is used tocontrol the system with its controller structure designed. A compensatinglearning control term is added to the error equation to deal with theuncertainties by evolutionary learning. Then, the computed torque method andthe GA learning control are combined to realize the perfect trajectorytracking. Simulation results show the efficiency of the control method.