Fault Detection and Diagnosis of Nonlinear processes Based on Kernel Principal Component Analysis

摘要：

A new fault detection and diagnosis method based on kernel principal component analysis (KPCA) is described.Firstly,kernel principal component analysis is adopted to detect fault. A new fault detection index is developed which replace Hotellings statistics T2 and SPE. The new index can simplify the fault detection task and avoid confusion.Secondly,a feature vector selection scheme is given to reduce the computation complexity of kernel matrix.Finally,kernel function gradient algorithm is used to diagnosis fault. The proposed method is applied to the simulation of Tennessee Eastman (TE) chemical process. The monitoring results confirm that the proposed method can effectively detect faults and diagnoses faults.

关键词： kernel principal component analysis fault detection fault diagnosis feature vector selection

作者: Jie XU Shou-song HU Zhong-yu SHEN

作者单位: College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing,210016, College of Automation Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing,210016, School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing,210042

会议类型: 国际会议

会议名称: 2009 International Conference on Information,Electronic and Computer Science(2009 国际信息、电子与计算机工程学术会议)

会议地点: 青岛

会议语种:英文

页码: 426-429

在线出版日期: 2009-11-21（万方平台首次上网日期，不代表论文的发表时间）