Fault diagnosis method based on nested iterative Fisher discriminant analysis

Abstract

The invention discloses a fault diagnosis method based on nested iterative Fisher discriminant analysis, which overcomes the problem of singularity of a within-class scatter matrix, the problem of number limit of discriminant components and a linear correlation problem of the discriminant components in the conventional method, fully explores potential information contained in process data and can be used for effectively differentiating the process data in different types. The method is simple and easy to implement, the performance of online fault diagnosis is greatly improved, the reliability and credibility of actual online fault diagnosis are enhanced, and faults can be accurately repaired by industrial engineers, so that safe and reliable operation of actual production and high quality pursuit of the product can be guaranteed.

Description

technical field [0001] The invention belongs to the field of chemical process statistical monitoring, in particular to a fault diagnosis method based on nested iterative Fisher discriminant analysis. Background technique [0002] As an important production method in industrial production, chemical process is closely related to people's life, and has been widely used in metallurgy, oil refining, papermaking, leather making and other fields. How to ensure the safety of chemical process production, improve product quality and economic benefits is the focus of attention. With the increasingly complex chemical process, online fault detection and diagnosis are becoming more and more important. Fault diagnosis refers to further judging what kind of fault has occurred after the fault is detected. Early diagnosis of faults can ensure the safe and reliable operation of production and the high quality of products, thereby avoiding major safety accidents, reducing casualties and impro...

AI Technical Summary

Problems solved by technology

However, when the online fault diagnosis method based on the traditional Fisher discriminant analysis is applied to the actual chemical process, there are three problems: first, the chemical process data is often highly coupled, which may lead to the singularity of the intra-class scatter matrix, so that it cannot Carry out singular value decomposition to extract potential information of process data

So far, do not see the research report relevant to the present invention yet

Images