Effective optimization for fuzzy model predictive control外语英文电子书.pdfVIP

IEEE TRANSACTIONS ON FUZZY SYSTEMS, VOL. 12, NO. 5, OCTOBER 2004 661 Effective Optimization for Fuzzy Model Predictive Control Stanimir Mollov, Robert Babuˇ ska, Janos Abonyi, and Henk B. Verbruggen Abstract— This paper addresses the optimization in fuzzy model of their ability to accurately approximate complex nonlinear predictive control. When the prediction model is a nonlinear fuzzy systems by using data combined with prior knowledge [2]–[4]. model, nonconvex, time-consuming optimization is necessary, Many successful applications of MPC using fuzzy models have with no guarantee of finding an optimal solution. A possible way around this problem is to linearize the fuzzy model at the current been reported [5]–[10] and recently several papers appeared in operating point and use linear predictive control (i.e., quadratic which different fuzzy MPC algorithms are analyzed and com- programming). For long-range predictive control, however, the pared; see [11]–[13]. The methods discussed in these papers can influence of the linearization error may significantly deteriorate generally be classified into two groups: 1) methods using di- the performance. In our approach, this is remedied by linearizing rectly the fuzzy model in the optimization procedure [7]–[9], the fuzzy model along the predicted input and output trajectories. One can further improve the model prediction by iteratively [12], [13], and 2) methods using a linearized model instead of applying the optimized control sequence to the fuzzy model and the fuzzy one [5], [6], [10], [12]. For example,