Model predictive control for a class of hybrid system based on linear programming

Aiming at the hybrid properties of an industrial process, predictive control based on a mixed logic dynamic model is presented to research a class of hybrid systems which comprises both logic and continuous controllers. The paper presents the algorithm for model predictive control based on 1-1 norm linear programming and the system's stability condition is also discussed. Simulation results show that the LP-based performance criterion has less computational time and cost than that of QP-based criteria. Thus the LP-based strategy is particularity well suited for solving control systems with large size and very stringent real-time requirement, which provides the complex industrial process with a new idea for the next generation.