基于主元分析的传感器故障检测与诊断  

Fault Detection and Diagnosis for Sensors Based On PAC

作　　者：胡暾[1] 赵哲身[1] 郑毅[1] 

机构地区：[1]上海大学机电工程与自动化学院,上海200072

出　　处：《仪表技术》2005年第6期30-32,共3页Instrumentation Technology

摘　　要：介绍主元分析方法在温度传感器故障检测与诊断中的应用。通过计算实际测量值和故障重构值及它们值的统计特性来识别故障类型;通过改进故障判别指数SPE及SVI,提高了故障检测与诊断的准确率。PCA（ principal component analysis） application for fault detection and diagnosis of temperature sensors was introduced. The fault type was distinguished by calculating the statistic characteristic of the measurement value, the reconstruction value and their difference, SPE and SVI were improved for higher accuracy.

关 键 词：温度传感器 故障检测与诊断 主元分析 平方预报误差 传感器有效指数 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]