基于PSO滚动优化的LS-SVM预测控制*

针对非线性时延系统、传统预测控制算法难以建立精确模型、控制精度不高的现状，提出一种基于最小二乘支持向量机（LS-SVM）的非线性系统预测控制算法。该算法通过LS-SVM对非线性系统输入输出数据序列的训练学习，建立其预测模型；然后运用粒子群(PSO)算法完成非线性预测控制的滚动优化。仿真结果表明，基于该方法的非线性系统预测控制具有较好的控制效果。

Predictive controlling of LS-SVM based on rolling optimization by PSO

It is difficult to build an accurate model using the traditional predictive control algorithm for the nonlinear time delay system, and the control accuracy is low. This paper proposed a nonlinear predictive control algorithm based on least squares support vector machines (LS-SVM) model. It obtained the nonlinear off-line model of the nonlinear system by LS-SVM to train a sequence data of input and output, and finished the whole rolling optimization procedure by PSO. The simulation results illustrate that the nonlinear predictive control using LS-SVM is effective.