Iterative Learning Control For Batch Process With Uncertainty

With the rapid development of modern industry, batch process has been paid more and more attention because of its characteristics of high added value and high precision. But compared with the traditional continuous process, the reaction mechanism of batch process is more complex, more susceptible to outside disturbs. In addition, batch process has higher production requirements, which makes control of batch process become a hotspot in the control community. This paper mainly discusses the application of iterative learning control algorithm in batch process with uncertainty.For linear uncertain batch process, this paper proposes a robust iterative learning control algorithm based on two degrees of freedom framework. According to the magnitude-frequency characteristic of the uncertain system, the robust modeling method is used to obtain the complete description of the system. The nominal model is used to design the feed-forward controller, while the feedback controller is obtained from the robust control scheme. By using the connection between the robust performance condition and the iterative learning control convergence condition, a learning controller is constructed straightforward from the existing 2DOF control system, which brings great convenience to the implementation of the system. It is shown that the control algorithm proposed in this paper can not only guarantee the robust stability of the first batch, but also the tracking error can converge to a small enough values in several batches.As to nonlinear batch process, it is difficult to use the traditional modeling and control methods to analyze the system. In this note, a moving window Gaussian process model is applied to finish the modeling of the system, a Batch to Batch control framework is presented based on the end point quality, which is predicted by a multi-step ahead method. The moving window Gaussian process model can adjust the parameters along the iteration axis, so that the proposed control scheme can guarantee good performance of the system under the condition of parameter drift.