| 基于主元空间动态模型的故障检测方法 |
| |
| 引用本文： | 文巧钧,宋执环,苗爱敏.基于主元空间动态模型的故障检测方法[J].南京航空航天大学学报,2011(Z1). |
| |
| 作者姓名： | 文巧钧 宋执环 苗爱敏 |
| |
| 作者单位： | 浙江大学工业控制技术国家重点实验室; |
| |
| 基金项目： | 国家高技术研究发展计划(“八六三”计划)(2009AA04Z154)资助项目; 浙江省自然科学基金(Y1080871)资助项目 |
| |
| 摘 要： | 建立了一种基于主成分分析的主元空间线性动态模型,将主成分分析的得分变量视为由高斯白噪声驱动的线性动态模型输出,可有效去除主元得分向量的动态相关性,动态模型参数可以通过期望最大化方法迭代辨识。分别在主元残差空间和主元空间线性动态模型上建立监控统计量,从而实现对动态过程的故障检测。通过数值仿真验证,该算法故障检测的检测率和误警率均表现良好。
|
| 关 键 词： | 故障检测 动态过程 主成分分析 状态空间模型 |
Fault Detection Approach Based on Linear Dynamical System in Principal Component Space |
| |
| Wen Qiaojun,Song Zhihuan,Miao Aimin.Fault Detection Approach Based on Linear Dynamical System in Principal Component Space[J].Journal of Nanjing University of Aeronautics & Astronautics,2011(Z1). |
| |
| Authors: | Wen Qiaojun Song Zhihuan Miao Aimin |
| |
| Institution: | Wen Qiaojun,Song Zhihuan,Miao Aimin(State Key Laboratory of Industrial Control Technology,Zhejiang University,Hangzhou,310027,China) |
| |
| Abstract: | To monitor dynamical process,a novel method based on principal component analysis is proposed.The method builds a linear dynamic system in the principal component space.Monitoring statistics are constructed in both residual space of principal component analysis(PCA) and linear dynamic system of principal component space.The proposed algorithm is implemented in monitoring a numerical dynamic process.Compared with other methods,the method shows its advantage in both false alarm rate and fault detection rate. |
| |
| Keywords: | fault detection dynamical process principal component analysis state space model |
| 本文献已被 CNKI 等数据库收录！ |
