An Adaptive Fuzzy Predictive Control Based on Support Vector Regression

Abstract

In this paper, an adaptive fuzzy Generalized Predictive Control (GPC) is proposed for nonlinear systems via Takagi-Sugeno system based Support Vector Regression (TS-SVR). The adaptive T-S fuzzy model is created using a support vector regression while the online learning procedure is obtained in two steps: first, the antecedent parameters of the TS-SVR are initialized using a k-means clustering and then iteratively adjusted using a back-propagation algorithm. Next, a sequential minimal optimization (SMO) algorithm is used to obtain the consequent parameters. Furthermore, the new TS fuzzy model is integrated into the GPC in order to control nonlinear systems. The performance of the proposed adaptive TS-SVR GPC controller is investigated by controlling the continuous stirred tank reactor (CSTR) system. The proposed TS-SVR GPChas shown good performance and efficiently controlled the nonlinear plant.