Data‐driven method for the quantitative fault diagnosability analysis of dynamic systems. Issue 8 (1st April 2019)

• Record Type: Journal Article
• Authors: Fu, Fangzhou; Wang, Dayi; Li, Linlin; Li, Wenbo; Wu, Zhigang
• Is Part Of: IET control theory & applications. Volume 13:Issue 8(2019)
• Page Start: 1197
• Page End: 1203
• Publication Date: 2019-04-01
• Subjects: sensor placement -- fault diagnosis; sensor placement -- direct identification methods -- fault diagnosability evaluation -- quantitative fault diagnosability analysis -- dynamic systems -- structural properties -- data‐driven evaluation method -- fault‐free data analysis; Control theory -- Periodicals; Automatic control -- Periodicals; 629.8312
• DOI: 10.1049/iet-cta.2018.5378
• Languages: English
• ISSNs: 1751-8644

Abstract

Abstract : The main objective of this work is to propose a method for the data‐driven evaluation of fault diagnosability in dynamic systems. This technique promises to be an efficient way to analyse the structural properties of unknown systems and can be used to optimise a system configuration, including sensor placement. The basic idea behind this method, which suitably integrates direct identification methods into the fault diagnosability evaluation, lies in the invariability of the characteristics of the dynamic behaviour corresponding to different fault cases. Data‐driven realisations of distinguishability based on both fault‐free data and fault data are proposed. Furthermore, a novel efficient method for computing distinguishability via the pseudoinverse is proposed. Finally, the effectiveness of the proposed method is illustrated via numerical and practical examples.