Research On The Algorithm Of Non-uniform Target Tracking Control

Iterative learning control is an approach to solve the exactly tracking problem for systems, which operate repetitively over a fixed time interval. It can improve the performance of controllers so as to produce zero error tracking during repetition of a command. However, in actual engineering, there are a large number of shortcomings which obstruct iterative learning control techniques from further applications. One of the main reasons is the target trajectory to be learned must be strictly identical throughout the repeated operations. No matter how small changes happened to the control objectives during the operation, you will have to resume from the beginning of the learning process. And the control knowledge achieved previously is no longer useful. This degrades the learning efficiency greatly.This paper has made a research on the non-uniform target tracking control problems from direct learning control and adaptive iterative learning control. Firstly, the concepts, basic ideas, contents, current researches and existent problems of iterative learning control are introduced.Chapter two introduces basic definitions and theorems to be used in this paper. In chapter three, the current researches on direct learning control are presented. Aiming at a class of linear systems with arbitrary relative degrees, the direct learning control problem of target trajectories is studied which have identical time scales and shapes but different magnitude scales, and the method on direct learning control is proposed. In chapter four, based on the investigation in chapter three, the direct learning control problems of target trajectories with different time scales and magnitude scales is further studied, the method to resolve the problems is also proposed. Lastly, the current researches on learning control based on Lyapunov methods are presented. In order to resolve the tracking control problems of non-uniform trajectories for a class of nonlinear systems with unknown relative degrees, a deep study is executed. The method of adaptive iterative learning control is proposed and its convergence is also proved.