Research On Iterative Learning Control For Harnessing The Nonrepetitive Disturbance

Iterative learning control (ILC) is a new intelligent control method developed in recent 20 years. In practical control of industrial process, some kinds of repeatable characteristics existed in many complex industrial processes. Current control information is revised by using the former control experience and the output errors of the system, the system output is caused to restrain as far as possible to the expect value according to the properties of industrial industry object. However the nonrepetitiveness always existed in the actual system and traditional iterative learning control algorithm cannot harness it well. The nonrepetitiveness is studied in this thesis. Firstly, the multi-loop method of iterative learning control is studied thoroughly, multi-loop algorithm of PID-type iterative learning control is workout by comparing Arimoto type multi-loop algorithm, the block diagram of multi-loop PID type ILC algorithm is mapped out and the convergence is proved, also. The method of iterative learning control in iterative domain is expanded by the multi-loop PID type ILC algorithm. Secondly, the internal model control is introduced, the framework of the iterative learning control based on internal model control is explored and the high-order ILC algorithm is obtained, also. Finally, the nonrepetitiveness is discussed, the traditional PID type ILC algorithm and internal model principle are combined by weighted method, then the PID-IMP type ILC algorithm are established and the parameters are optimized. The validity of this algorithm on unstable system is also explored. Simulation examples are presented to illustrate the effectiveness of the proposed method. The nonrepetitiveness can be harnessed by the algorithm and the application field of iterative learning control will be broaden, iterative learning control will be better applied into actual production by the new iterative learning control algorithm.