Application Study Of Multi-Step Robust Model Predictive Controller In Beer Fermentation Temperature Control

Robust Model Predictive Control(RMPC) as a sort of advanced control theory, has been an extensive attention by industrial and theoretical horizon, due to its adaptability for the complex industrial processes, and good control performance. Over the past decade, model robust predictive control has been rapidly developed and many important new algorithms were proposed. But according to the practical industrial objective with large-time delay, uncertainty and time varying, how to better handle with many other issues such as control performance, online computation burden and feasible region of a RMPC controller, which are important in practical applications, are still expected to be further studied.In this paper, according to the engineering background of beer fermentation process, a new robust model predictive control online algorithm which based on invariant set theory and multi-step control sets method is proposed, and is applied it to the beer ferment temperature process which has large phase lags and uncertainty. The simulation results show the algorithm's validity and practicability in controlling beer ferment temperature.The main contents of the research in this paper are as follows:(1) For the practical industrial process, the input increment as a new input, then an augmented model with polytopic uncertainty and large phase lags was built. Based on this model, robust model predictive control algorithm with multi-step control sets was studied. According to the characteristics of augmented model's input matrix, by making use of the gain scheduling and multi-step control set, robust predictive control algorithm with the parameter-dependent feedback law is proposed. The simulation results show the advantages and disadvantages of two algorithms,and the robust model predictive control algorithm with gain scheduling and multi-step control set can improve the control performance of the system better than pure application multi-step control set.(2) The beer ferment production technology and control requirements are elaborated in detail. Emphasis is on introduce a class of large delay, uncertainty and time varying beer fermentation temperature plants. The diffculty and control in brewage production proeess was analysed. And discrete-time model with polytopic uncertainty of the beer temperature control system is established, through mechanism analysis on the model structure and through the test to identify the parameter values of model structure. Finally the RMPC controller was designed by using the proposed algorithm, and the discrete-time model for temperature process was analysed, the simulation results illustrate that by this controller, temperature inside beer ferment tank is good tracked the technology cure.