Multi-feature fault diagnosis method based on negative evidence reliability structure

Abstract

The invention discloses a multi-feature fault diagnosis method based on a negative evidence reliability structure. The multi-feature fault diagnosis method comprises the following steps: step 1, generating triangular fuzzy numbers according to a fault sample data set and to-be-tested sample data respectively; 2, generating a negative evidence reliability structure of each fault feature according to triangular fuzzy number matching; 3, generating an evidence reliability structure of each fault feature; 4, generating the weight of each fault feature according to the fault sample data set; and 5,fusing the evidence reliability structure of each fault feature and judging the fault type to which the to-be-tested sample belongs. On the basis of the triangular fuzzy number, in combination with the advantage of evidence theory processing uncertainty, the negative evidence reliability structure of each fault feature is generated according to triangular fuzzy number matching, the evidence reliability structure is generated for the negative evidence reliability structure, the weight of each fault feature is calculated, and the evidence reliability structure of each fault feature is fused, sothat the fault type of the sample is identified; uncertain information can be effectively processed, and the fault diagnosis accuracy is improved through fusion.

Description

technical field [0001] The invention belongs to the field of fault diagnosis, and in particular relates to a multi-feature fault diagnosis method based on negative evidence reliability structure. Background technique [0002] With the improvement of modern science and technology and production level, automation and intelligence are gradually applied in more and more fields, such as industry, scientific research, and medicine. In actual use, the safety of system operation is a crucial issue, and timely fault diagnosis of the system can effectively detect its operating status and avoid safety accidents. Therefore, the research on fault diagnosis technology has very important practical significance. [0003] Fault diagnosis mainly studies how to effectively identify and locate the system operating status and fault types. Existing fault diagnosis methods include model-based methods, signal-processing-based methods, data-driven methods, and knowledge-based methods. As modern e...

AI Technical Summary

Problems solved by technology

In addition, due to factors such as environmental noise, system noise, and measurement errors, fault information is generally uncertain or even conflicting, which may lead to large errors in fault diagnosis results.

Images