Iterative Learning Control of Exponential Variable Gain Based on Initial State Learning for Upper Limb Rehabilitation Robot

Electrical and Control Engineering, Lanzhou, China Trajectory tracking problem for a nonlinear system of upper limb rehabilitation robot over a finite time interval, an exponential variable gain D-type Iterative Learning Control (ILC) law with initial variable learning is designed. The method uses closed-loop ILC with exponential variable gain for both the control input and the the initial value of the system. Based on the operator theory, the convergence of the system with arbitrary initial state under the ILC is strictly proved. At the same time, the sufficient conditions for the convergence spectrum radius form of the ILC method are given. Compared with the ILC of fixed gain, the control method not only accelerates the convergence speed, but also solves the problem that the ILC requires strict repetition of the start state. Finally, the control effect of this method is verified by experimental simulation, the application of the control method in trajectory tracking of rehabilitation robot has achieved good control effect..