Delay Batch Process The Pd-type Iterative Learning Control

More and more attentions are paid to control of batch processes with the extensive use of this process in modern industry. But the control of batch processes is a very challenging problem because of its own strong delay, nonlinear, time-varying parameters and model uncertainty, etc. This dissertation mainly studies nonlinear batch process with initial deviation, time delay and repetitive output disturbance based on PD-type iterative learning control algorithm.The main contents are included as follows:(1)This dissertation discusses the convergence of iterative learning algorithm in three cases according to time delay:the closed-loop PD-type learning algorithm with forgetting factor is given for a class of nonlinear systems with state delay, the advanced open-loop PD-type learning algorithm with forgetting factor is given for a class of nonlinear systems with control delay, the advanced open-closed-loop PD-type learning algorithm is given for a class of nonlinear systems with state delay and control delay. The algorithm's convergence condition of spectral radius form is given respectively. Then the convergence of algorithm is proved based on operator theory andλnorm.(2) The consistency of the learning algorithm and the convergence condition is proved through the simulation. The results of the simulation between above algorithms and ordinary open-loop PD-type learning algorithm prove the superiority of the method in trajectory tracking and disturbance suppression.(3)The convergence rate of closed-loop PD-type and open-closed-loop PD-type learning algorithms are studied. The factors that affect the convergence rate are obtained in theory, which provides the theoretical basis for the further improvement of convergence rate.(4) In the context of the production process of Unsaturated Polyester Resin--a kind of typical batch processes,The mathematical model has been constructed against Continuous Stirred Tank Reactor (CSTR for short) which is the key control object in the production process.The closed-loop PD-type learning algorithm and PID algorithm are applied to simulation for this model.The results of simulation prove that ILC algorithm can obtain a better control effect.