Research On The Human-Robot Cooperative Control Of A Power-Assist, Upper Extremity Exoskeleton System Driven By Pneumatic Muscles

As a kind of wearable robot systems, the power-assist upper extremity exoskeleton has been widely studied for its ability to enhance the user’s function. Exoskeleton driven by pneumatic muscles (PM) has good flexibility, but its dynamics and the pneumatics system’s inherent nonlinearity make the system harder to control. This dissertation is focused on the research about modeling and human-robot cooperative control of an upper extremity exoskeleton driven by pneumatic muscles.Firstly, PM’s mechanical properties are tested in order to establish the experimental model. Based on this model, two kinds of discrete sliding mode controllers are designed for PM test system. Furthermore, a single-degree-of-freedom system’s continuous sliding mode controller is designed, and the experiment results prove that the sliding mode control law has better robustness.Secondly, the double-degree-of-freedom mechanical arm’s coupled dynamics is analyzed and the exoskeleton’s state equations are deduced next. In order to study sliding mode law’s decoupling performance, the arm end’s tracking controller is designed and it’s verified by experiments.Thirdly, the human-robot interaction (HRI) model is established and the reference signals produced by the contact force are used to judge the wearer’s moving intentions accurately. The model’s good anti-interference performance and the controller’s quickness are proved in the experiments.Finally, we developed the PC-based software with measurement and control functions for different control objectives. Besides, the hardware system is set up with FPGA as the PWM control unit.