Fault Diagnosis of Rotating Machinery Based on Dimensionless Index and Two-sample Distribution Test

Nai-quan Su ； Xiao Li ； Qing-hua Zhang ； Chong-lin Huang

《電腦學刊》 31卷3期 (2020/06) Pp. 1-10

https://doi.org/10.3966/199115992020063103001

摘要

Dimensionless indexes are used to identify the faults type of rotating machinery, there is an overlap between the dimensionless index ranges of each fault, and result in poor fault diagnosis. So in this paper, an algorithm is proposed a dimensionless indexes and two-sample distribution test. This method are compared the cumulative probability distribution function between the known fault samples and the tested fault samples, and find out the maximum statistical distance. According to the similar degree of the two samples, and the tested fault sample is able to be diagnosed. The method is applied in the simulation experiment of petrochemical unit show that the method is simple and effective, and it has the strong fault identification ability to extract the fault feature accurately. This provides the guidance for rotating machinery fault diagnosis and other large equipment.

關鍵字

dimensionless indexes ； experimental verification ； fault diagnosis ； two-sample distribution test