摘要
一般辨识方法在传感器间进行预测来对传感器故障诊断时 ,其准确的冗余关系难以找到 ,针对这一缺点 ,提出了一种用主元分析技术建立传感器间的预测模型的方法。它同时克服了对传感器测量的数据直接进行主元分析需要诊断的传感器较多的不足。最后 。
It is difficult to determine the redundant state using an identification algorithm in sensor fault diagnosis. A new approach is presented to build a prediction model between sensors based on Principle Component Analysis (PCA). It overcomes the defect of demanding more sensors when testing the data of sensors. A simulation experiment testifies to the effectiveness of this method.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2002年第2期57-59,共3页
Journal of Wuhan University of Technology：Information & Management Engineering
关键词
主元分析
传感器
故障诊断
Principle Component Analysis
sensor
fault diagnosis
