Research On Algorithm Of Iterative Learning Model Predictive Control

In order to achieve trajectory tracking for a class of repetitive system with non-repetitive disturbances, traditional methods could not achieve tracking of desired trajectory. An iterative learning control algorithm based on model predictive control is studied in this paper. First of all, the principle and algorithm of iterative learning control and model predictive control are introduced. As a single algorithm can not tracking the desired trajectory perfectly, combination of the two algorithms is necessary for Linear system. Based on the temporal model of tracking error, the model predictive control is applied to improve the tracking proformance along the time index to control the current stochastic disturbances. In order to restrain the repetitive disturbances, D-type Learning law is applied along the iterative index. The simulation analysis of this algorithm is completed.A class of nonlinear systems could be describe by T-S fuzzy model. An Iterative Learning Control combine with model predictive control by using T-S fuzzy model is proposed for nonliner system. Based on fuzzy rules, a nonlinear system is divided into several linear subsystems.Then the algorithm is applied to each linear subsystem. Temporal state space form is used along time index, an iterative learning law of T-S fuzzy model is given along at iterative index. On the base of the state-space model, The combination of two control methods law is given. Finally, Simulation results shows that with the number of iterations grow, it can be perfect tracking of desired trajectory by using this algorithm.