一种基于2次核函数支持向量机的多步预测控制算法  

A Predictive Control Algorithm for Nonlinear Systems Based on SVM with Quadratic Polynomial Kernel Function

作　　者：钟伟民[1] 钱锋[1] 皮道映[2] 

机构地区：[1]华东理工大学自动化研究所,上海200237 [2]浙江大学工业控制国家重点实验室,杭州310027

出　　处：《华东理工大学学报（自然科学版）》2007年第4期524-528,共5页Journal of East China University of Science and Technology

基　　金：国家973计划(2002CB3122000);上海市自然科学基金(05ZR14038);上海市科委科技攻关项目(04DZ11010;05DZ11C02);上海市科委重大基础研究(05DJ14002)

摘　　要：提出了一种基于2次多项式核函数支持向量机的多步预测控制方法。通过黑箱辨识和线性化技术得到非线性系统的近似模型,根据预测控制机理,最小化滚动时域的二次型目标函数,利用模型算法控制的方法得到控制器的解析输出。通过一个标准预测模型和一个工业用连续搅拌槽式反应器的模型仿真验证了该控制器的性能,仿真结果表明:该控制器有着良好的预测性能。In this paper, a support vector machine （SVM） with quadratic polynomial kernel function based nonlinear model multi-step-ahead predictive controller is studied. An SVM based predictive model is established using black-box identification techniques and linearization is done to get the approximate linear model. Under the quadratic objective functions with receding horizon, the controller outputs can be obtained using the method of model algorithmic control. The effect of the controller is demonstrated on a recognized benchmark problem and a continuous-stirred tank reactor. Simulation results show that multistep-ahead predictive controller can be well applied to nonlinear systems, with better performance in following reference trajectory and disturbance-rejection.

关 键 词：非线性模型预测控制 支持向量机 模型算法控制 系统辨识 核函数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]