机构地区: 广东工业大学自动化学院
出 处: 《广东工业大学学报》 2007年第2期76-79,共4页
摘 要: 针对现今在非线性系统盲辨识研究中遇到的困难,提出了一种基于最小二乘支持向量机(LS-SVM)的非线性系统盲辨识方法.该方法直接对非线性系统输出进行过采样,运用LS-SVM非线性建模技术,并结合输入的分布特性,从而完成非线性系统的盲辨识.介绍了盲系统辨识问题的研究内容及过采样技术原理,对LS-SVM的盲系统辨识机理和算法步骤进行了阐述.仿真结果表明了该方法在解决非线性系统盲辨识问题上的切实可行性. To solve the nonlinear problems in blind system identification, a novel blind nonlinear system identification approach based on Least-Square Support Vector Machines (LS-SVM) is investigated in this paper. By oversampling the system outputs, more information of the system characteristics can be observed to blindly identify nonlinear systems. The LS-SVM based mathematical approximation provides an adequate modeling of the unknown sys- tem given the distribution knowledge of the system inputs. The paper describes the mechanism of the oversampling and LS-SVM, explains the procedure of the algorithm of the blind system identification. Simulation results demonstrate the effectiveness of this approach.
领 域: [自动化与计算机技术] [自动化与计算机技术]