基于SVM和PSO的新型非线性模型预测控制  

New NMPC Based on the Integration of SVM and PSO

作　　者：李涛[1] 吕勇哉[1] 陈鹏[1] 

机构地区：[1]上海交通大学自动化系,上海200240

出　　处：《控制工程》2008年第S2期102-105,共4页Control Engineering of China

摘　　要：对于具有强非线性或复杂非线性约束的系统,通过非线性模型的线性化和二次规划优化实现非线性模型预测控制,难以取得满意的结果。提出了一种基于支持向量机模型和粒子群优化的非线性模型预测控制系统的算法。仿真实例表明了支持向量机模型的泛化能力和粒子群优化的寻优速度及能力,证明了将其运用于非线性模型预测控制中的可行性。To the high nonlinear and complex constrained system,it is difficult to obtain satisfying performance in model predictive control using of nonlinearity of nonlinear model and quadratic programming(QP)optimization. An algorithm of nonlinear model predictive control system based on support vector model and particle swarm optimization is proposed.The simulation result shows the generalization ability and optimization speed and ability of particle swarm optimization,and the feasibility to use it in nonlinear mo...

关 键 词：非线性模型预测控制 粒子群优化 支持向量机 滚动优化 收敛性 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]