华侨大学学报（自然科学版）2012，Vol.33Issue(6)：601-607,7.
基于MSPM的故障诊断技术研究现状与展望
Research Status and Prospect of Fault Diagnosis Technology Based on MSPM
摘要
Abstract
First of all, this paper introduces the basic ideas and applications of statistical process monitoring method based on principal component analysis (PCA) model? Partial least squares (PLS) model and independent component analysis (ICA) model. The present research situation and development trend about various methods are reviewed. Secondly, by combining fault prediction technology with the traditional statistical process monitoring technology, fault prediction method based on multivariate statistical process monitoring (MSPM) can be realized. And some research results are also introduced. Finally, six difficult problems in multivariate failure prediction technology such as non-Gaussian, non-linear, multi-modal, probability distribution, intermittent process fault prediction and application verification are discussed respectively.关键词
多元统计过程监控/故障诊断/故障预测/主元分析/偏最小二乘法/独立分量分析Key words
multivariate statistical process monitoring/fault diagnosis/ fault prediction/ principal component analysis/ partial least squares / independent component analysis分类
信息技术与安全科学引用本文复制引用
马洁,党爱民,李刚,周东华..基于MSPM的故障诊断技术研究现状与展望[J].华侨大学学报（自然科学版）,2012,33(6):601-607,7.基金项目
国家自然科学基金资助项目(61273173,61028010,61021063) （61273173,61028010,61021063）
北京市自然科学基金资助项目(4122029) （4122029）
