Point-to-Point Iterative Learning Control Based on Updating Reference Trajectory with Constrained Input

Abstract

The point-to-point tracking control method under constrained input is proposed by using updating-reference and an integrated predictive iterative learning control strategy. A reference trajectory through the desired key points is adopted and updated batch-to-batch, and then the whole system is described as 2D model. By using the integrated predictive ILC, the control method can depress effectively disturbances. For the constrained input, its convex set is abstracted and the procedure of calculating the constrained input is presented in detail. Comparing with gradient based point-to-point control algorithms, updating- reference relaxes the output constraints and the proposed algorithm can lead to faster convergence. Simulation results of a numerical model have demonstrated the effectiveness of the proposed method.