Robust Controller Design for the Cooperative Control of Lower Extremity Rehabilitation Robot

Abstract

In this paper, a robust controller design approach is proposed for Lower Extremity Rehabilitation Robot (LERR) with automatic gait adaption. Based on the developed LERR dynamic model, the robot system is first decoupled into some independent second-order integral systems using inverbility decoupling method. Then a robust optimal PID controller is proposed for each sub-system to cope with the uncertainty in the system modeling by converting the PID controller design problem into a static output feedback H 2 / H ∞ multiobjective optimization problem. The adaptive control methods are further investigated for a compliant gait trajectory planning. The performance of the designed controller is finally verified in the Adams-MATLAB/Simulink environment to show the effectiveness of the proposed robust control approach. © 2014 SERSC.