摘要
基于主元分析的传感器故障检测方法是一种以数据统计为依据的重要故障检测方法。已有的研究一直忽视了对HotellingT2和HawkinsT2H指标在传感器故障检测中的意义和作用的讨论。通过分析T2和T2H指标在主元分析的热力系统传感器故障检测中的不同适用性,提出了应用T2H的故障检测方法。该方法克服了T2统计参数对传感器小故障不敏感的问题。仿真实验的结果验证了该结论。
Sensor fault detection, based on principal component analysis, is an important way for detecting sensor faults by statistics. Past research has neglected, up to now, to take notice of discussions concerning the importance and functioning of Hotelling' s T^2 and Hawkins' TH^2 indices in the detection of sensor faults. Based on an analysis of the different applicability of these two indices in the detection of sensor faults in thermal systems, a relevant fault detection method is being proposed, which solves the problem of the statistical parameter T^2 ' s in sensitivity to minor sensor faults. Simulation test results have vindicated this conclusion.
出处
《动力工程》
CSCD
北大核心
2007年第3期376-380,共5页
Power Engineering
关键词
自动控制技术
主元分析
故障检测
传感器
automatic control technique
principal component analysis
fault detection
sensor