A generalized model predictive control method for series elastic actuator driven exoskeleton robots

In order to optimize the control performance of the compliant actuator system, a generalized predictive control (GPC) method is introduced in the controller design. Compared with conventional predictive control methods which largely depend on the modeling, the proposed GPC method is almost model-free. The stability of closed-loop system has been rigidly proved. Various experimental tests of a series elastic actuator driven exoskeleton robot are verified the effectiveness of the proposed methods.

Automated summary

The control performance of the compliant actuator system was optimized using a generalized predictive control method, which is almost model-free and has proven system stability. Experimental tests on a series elastic actuator driven exoskeleton robot confirmed the effectiveness of the proposed methods.