This paper describes the application of system identification techniques and robust control strategies to a pneumatic muscle actuator system. Due to the inherent nonlinear and time-varying characteristics of this system, it is difficult to achieve excellent performance using conventional control methods. Therefore, we apply identification techniques to model the system as linear transfer functions and regard the un-modeled dynamics as system uncertainties. Because robust control is well-known for its capability in dealing with system uncertainties, we then apply robust control strategies to guarantee system stability and performance for the system. This work is carried out in three parts. First, the pneumatic muscle actuator system was modeled as linear transfer functions. Second, robust control theorem were utilized to design a Hinf robust controller to deal with system uncertainties and performance requirements. Finally, the designed controller was implemented for experimental verifications and compared with a conventional PID controller. From the experimental results, the proposed Hinf robust controller is deemed effective.
Nonlinear Control of a Pneumatic Muscle Actuator System