摘要
支持向量机 (SVM)是一种新的通用学习机器 ,它从结构风险最小化的角度 ,分析了学习过程的一致性、收敛速度等。SVM能以任意精度逼近一类函数 ,而与输入的维数无关 ,克服了传统神经网络用于系统辨识的维数灾问题及结构难以确定等缺点。基于这一特性研究了对非线性动态系统的辨识问题 ,仿真结果表明SVM用于系统辨识有良好的辨识效果 。
Support vector machine (SVM) is a new general learning machine,which analyzes the consistency of learning and speed of convergence from structure risk minimization principle.SVM can approximate any function at any accuracy independent of dimension of input,which overcomes the problem of dimension curse and hard decision of structure that traditional neural networks have met in system identification.Based on the characteristics of SVM the application of SVM for nonlinear dynamic system identification has been studied.Simulation shows the good identification result,and the future research direction is pointed out.
出处
《测控技术》
CSCD
2002年第11期54-56,58,共4页
Measurement & Control Technology
