Optimizing The Boundary Value Problem Based On The Arithmetic Of Iterative Learning Control

Iterative learning control is a control technique to improve repetitive motion process, mechanics, equipment or system’s transient response and tracking. It modifies incorrect control signals form the deviation between the actual output and desired output, so as to improve the tracking performance of the controlled system.This dissertation introduces the ILC at first, including principle, direction of research and applications both in and out of the country. Moreover, it introduces the initial states problem briefly. This dissertation studies the P type iterative learning control system to obtain better tracking performance on error’s rapid, monotone convergence in the iteration domain. In order to improve the control precision of the controlled system and accelerate the convergence speed of iteration domain, the initial input values of the boundary conditions is optimized, and the convergence is proved by simulation analysisIn order to improve the convergence properties of the iterative learning control algorithm, the study use the genetic algorithms to optimize the iterative learning control algorithm, on the basis of the traditional PD type iterative learning control algorithm, the gain parameters of iterative learning control are optimized, in order to obtain the optimized iterative learning law. Then, by further research on optimizing the boundary value problem of iterative learning control, the influence of different boundary value on the performance of the system has been overcome. This algorithm guarantees algorithm has the properties of fast convergence, at the same time, it further expand the convergence condition of the algorithm.Based on genetic algorithm, the dissertation makes a research on boundary value problem of iterative learning control. By comparing with the traditional PD type iterative learning control algorithm, the conclusion is that the iterative learning control based on genetic algorithm can make the controlled system achieve a total track of the desired trajectory at the same time to improve the convergence speed. In actual application, based on genetic algorithm, an optimization algorithm of iterative learning control’s boundary value also has a good application prospect.