Relevant parameter fault classification method based on big data fusion clustering analysis

Abstract

The invention provides a relevant parameter fault classification method based on big data fusion clustering analysis. By use of the fault classification method, fault data is picked up according to an interpretation rule from mass data which operates in equipment, supervised machine independent clustering is carried out, an automatic classification result of relevant parameter faults is formed, the problem that existing equipment fault diagnosis excessively depends on an expert knowledge base but ignores an incidence relation among deep nonlinear coupling parameters of each subsystem is solved, and the problem that mass effective data can not be favorably mined and utilized in practical equipment type operation is solved. Meanwhile, since the implementation of the fault classification method does not need to depend on the accurate physical modeling of object equipment, a difficulty that a traditional complex system is difficult in modeling is avoided, the intelligent classification of faults and relevant parameter analysis on the basis of mass data mining can be realized, and the method has a fault classification capability with controllable accuracy.

Description

technical field [0001] The invention relates to the field of equipment failure prediction and health management (PHM), in particular to a method for classification of associated parameter failures based on big data fusion cluster analysis. Background technique [0002] Fault prediction and health management have developed into an important supporting technology and foundation for system logistics support, maintenance and autonomous health management in the aerospace field. In the "National Medium- and Long-Term Science and Technology Development "Life prediction technology" is proposed as a cutting-edge technology. In the development reports of aerospace and aviation science and technology disciplines in recent years, PHM technology is listed as a key and supporting technology. [0003] PHM technology has become an interdisciplinary and popular research direction covering basic materials, mechanical structures, energy, electronics, automatic testing, reliability, information...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to solve the technical problem that the existing data-driven PHM method has difficulty in obtaining fault data. The status quo that equipment operation data containing massive information has not been effectively mined and utilized

Images