收录:
摘要:
Principal component analysis (PCA) is a common fault detection method. But it is difficult to get high accuracy, if it is applied to complex nonlinear system. Faced with complex system such as chiller, this paper proposes using kernel principal component analysis (KPCA) for fault detection. But, the selection of kernel parameters is a problem in the implement of KPCA algorithm. Genetic algorithm (GA) is used to determine the kernel parameter through minimizing false alarm rate and maximizing detection rate. This method is verified by ASHRAE 1043-RP data. The results show that it is better than PCA. And it can improve the accuracy of fault detection.
关键词:
通讯作者信息:
电子邮件地址:
来源 :
PROCEEDINGS OF THE 28TH CHINESE CONTROL AND DECISION CONFERENCE (2016 CCDC)
ISSN: 1948-9439
年份: 2016
页码: 2951-2955
语种: 英文
归属院系: