T-S模型的遗传算法和支持向量机辨识  

T-S Model Identification Based on Genetic Algorithm and Support Vector Machine

作　　者：丁学明[1] 樊海军[1] 

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《控制工程》2009年第5期591-593,597,共4页Control Engineering of China

基　　金：上海高校选拔培养优秀青年老师科研专项基金资助项目(21020);上海理工大学博士科研启动基金资助项目(X641)

摘　　要：基于非线性系统的输入输出数据,辩识对象的T-S模型。提出基于遗传算法和最小二乘支持向量机的辨识方法,利用遗传算法聚类进行结构辨识,每个类代表一条规则,规则数等于类数量,类中心作为该规则的隶属度函数中心类数;同时考虑模型辨识精度,实现全局优化;参数辨识采用基于结构风险最小化的最小二乘支持向量机方法,综合考虑模型复杂度和辨识误差。仿真结果证明了算法的有效性,辨识精度高,泛化能力强。T-S model identification is studied for nonlinear plant based on its input-output data. An identification method is presented based on genetic algorithm （GA） and least square support vector machine （LSSVM）. The GA clustering method is employed in the structure identification. Each cluster is equal to a rule, and the cluster center is regarded as center parameter of gauss MF. The global optimization is realized by considing identification precision. The parameter identification is depended on LSSVM based on structure risk minimum. which considers model complexity and identification error. The simulation result shows the effectiveness of the proposed method with high identification precision and good generalization.

关 键 词：T—S模型 结构辨识 参数辨识 遗传算法 最小二乘支持向量机 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]