振动发电机的支持向量机建模研究  

Research on SVM Modeling of Linear Reciprocating Generator

作　　者：贺超英[1] 王少喻[2] 

机构地区：[1]中南林业科技大学计算机与信息工程学院,湖南长沙410004 [2]湖南工程职业技术学院土木工程系,湖南长沙410151

出　　处：《自动化仪表》2011年第11期9-12,共4页Process Automation Instrumentation

基　　金：中南林业科技大学青年科学基金资助项目(编号:2008032B)

摘　　要：研究可直接利用物理机械振动发电的直线振动发电机对节能环保具有重要的意义。采用具有准确实时性的支持向量机(SVM)建模方法对新型直线振动发电机进行研究。首先建立振动发电机的有限元模型,以获得表征电机性能参数的样本数据集;然后运用SVM回归建模方法对此样本数据集进行学习,建立直线发电机的非线性模型;最后将有限元模型(FEM)和SVM模型进行比较。仿真表明,SVM模型可靠高效,对实现直线振动发电机的实时在线控制及在线预测具有非常重要的应用价值。It＇s significant to research the value of the linear reciprocating generator that is generating electric power by directly adopting physical and mechanical vibration to the energy saving and environment contribution.The new type of linear reciprocating generator is studied by using the support vector machine（SVM） modeling method that features precisely real-time performance.Firstly,the finite element model（FEM） of the generator is established to obtain sample data set indicating performance parameters of the generator;then the sample data set is learnt with the support vector machine regression modeling method to setup nonlinear model of the generator;finally,the FEM and SVM model are compared.The simulation shows that the SVM model is highly reliable,it offers very important applicable values to implement real-time online control and online prediction for linear reciprocating generator.

关 键 词：直线振动发电机 有限元模型 支持向量机 感应电动势 非线性 

分 类 号：TM359.4[电气工程—电机]