Synthesis of Real-Time-Feedback-Based 2D Iterative Learning Control–Model Predictive Control for Constrained Batch Processes with Unknown Input Nonlinearity

Dewei Li,Yugeng Xi,Jingyi Lu,Furong Gao
DOI: https://doi.org/10.1021/acs.iecr.6b03275
2016-01-01
Industrial & Engineering Chemistry Research
Abstract:Batch process is an important category of industrial processes. Recently, the combination of the real-time-feedback-based iterative learning control (ILC) and the model predictive controller (MPC) has demonstrated its advantage when applied to the batch process. In practical applications, the plants are always nonlinear and the model cannot be known exactly. Therefore, how to design such a control strategy for the constrained batch processes with unknown input nonlinearities and guarantee the convergence is interesting and valuable. Inspired by the dual mode MPC, this paper proposes a two-mode framework for the constrained ILC MPC to solve this problem, which is constructed by a real-time-feedback-based strategy followed by a run-to-run strategy. Under the proposed framework, an invariant updating strategy based on the run-to-run strategy is developed for constrained batch processes to act as the second mode, which describes the situation of infinite batch and gives an estimation on the upper bound of the influence from the nonlinearity. Then, a real-time-feedback-based ILC MPC with a two-dimensional (2D) model is designed for the considered batch process, which adopts the time-varying horizon, and is proven to be recursively feasible and convergent. The case studies verify the results of the paper.
What problem does this paper attempt to address?