LS-SVM based stable generalized predictive control

A stable generalized predictive controller (SGPC) based on least squares support vector machine(LS-SVM) model is presented. For a given closed-loop control configuration, a receding horizon optimal control law which can stably track the constant set point free-offset is obtained by optimizing the objective function over the future reference signal. The system model is built by using LS-SVM, and then the obtained model is transformed into input-output relation of the controlled system, which can be employed by SGPC directly. The detailed implementation of the presented algorithm is given and the algorithm is applied into a pulp washing process. The simulation results revealed the effectiveness and merit of the presented algorithms.