Model Predictive Control For Piecewise Linear System Based On Mixed Logical Dynamic Model

Due to the rapid development of industry and technology, for the sake of meeting more precise modeling and control requirement the classic controller design method based on"Piecewise Linear model"rather than models either continuous models or discrete models has been come up. Piecewise linear systems are used to describe one class of system with both continuous process and discrete process. To certain extent, large number of industrial processes can be considered as a typical piecewise linear system. Meanwhile model predictive control, which acts as a typical control design method with the merits of handling the constraints and uncertainty flexibly, has been widely used since the day it was put forward. Above all, doing research on model predictive control for piecewise linear systems are of great practical importance.This paper takes the typical model of large-domain linear systems as research object, with the help of mixed logical dynamic model, to take an investigation on model predictive control of dynamic systems. Through logic variables, piecewise linear systems can be modified into a single-linear model and affiliated constraints with mixed logical dynamic systems. The main dedication of this paper includes:1. This paper takes a detailed review over piecewise linear systems based on mixed logical dynamics. Based on the structure of mixed logic dynamics, the basic controller frame for piecewise linear systems are mentioned. Furthermore, referring to model predictive control, the detailed formulation of predictive controller design and simulation for piecewise linear systems are given.2. For the industrial process this paper focuses on the satisfying optimal predictive control of switching multiple systems based on mixed logic dynamic model with the reference of workers'experience. The method makes a balance of objectives, constraints and satisfying degree of the workers through fuzzy decision. Meanwhile through fuzzification of the constraints and objectives the problem of meeting the performance requirements have been turned into that of reaching the highest satisfying degree. The simulation example verifies the feasibility of the method.