| 基于LAR和在线LS-SVR的非线性时间序列故障预报 |
| |
| 引用本文： | 苏圣超,陈国明,胡义刚.基于LAR和在线LS-SVR的非线性时间序列故障预报[J].昆明理工大学学报(理工版),2010,35(2):66-71. |
| |
| 作者姓名： | 苏圣超 陈国明 胡义刚 |
| |
| 作者单位： | 上海工程技术大学,工程实训中心,上海,201620 |
| |
| 基金项目： | 上海市高校优秀青年教师专项基金资助项目(项目编号:gcd07047) |
| |
| 摘 要： | 针对非线性系统的故障预报,设计了一种在线最小二乘支持向量回归机(LS-SVR)算法,提出了一种基于在线LS-SVR和线性AR(LAR)混合预测的故障预报新方法.用LAR对非线性系统进行局部线性建模,用LS-SVR在线补偿局部线性模型的建模误差,实现了非线性时间序列的一步预测,并推广到N步预测.基于已知的正常时间序列数据,直接对当前N步预测值进行异常估计,实现故障预报,提高了实时性.同时方法的误检率和漏检率还可人为调整,对不同对象具有普遍性.仿真实验证明了方法的有效性.
|
| 关 键 词： | 故障预报 时间序列预测 异常估计 最小二乘支持向量回归机 |
Nonlinear Time Series Fault Prediction Based on LAR and Online LS-SVR |
| |
| SU Sheng-chao,CHEN Guo-ming,HU Yi-gang.Nonlinear Time Series Fault Prediction Based on LAR and Online LS-SVR[J].Journal of Kunming University of Science and Technology(Natural Science Edition),2010,35(2):66-71. |
| |
| Authors: | SU Sheng-chao CHEN Guo-ming HU Yi-gang |
| |
| Affiliation: | Industrial Engineering Training Centre/a>;Shanghai University of Engineering Science/a>;Shanghai 201620/a>;China |
| |
| Abstract: | A novel fault prediction method for the nonlinear systems is presented in this paper,which is based on the combined online least square support vector regression(LS-SVR) and autoregressive(AR) models.The AR model is used to fit the linear part of time series.The nonlinear part and the approximate error are compensated by the online least square support.A one-step-ahead prediction method of the nonlinear time series is thus achieved,and it is then extended to N-step-ahead prediction.Based on the normal time ... |
| |
| Keywords: | fault prediction time series prediction abnormity estimation LS-SVR |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
