A composite fault diagnosis method for rolling bearings embedded in fault semantic space

Abstract

A rolling bearing composite fault diagnosis method embedded in the fault semantic space, first obtain the vibration signal samples of the rolling bearing in each healthy state, use the feature extraction and dimensionality reduction module to automatically obtain the characteristics of the vibration signal samples and reduce the dimensionality; then use Gauss-Bernou Use the restricted Boltzmann machine to calculate the energy function value of the sample, and distinguish whether the sample belongs to the composite fault; finally, through the multivariate classifier and the linear supervised automatic encoding machine, the health status identification of the single fault sample and the composite fault is completed ; The present invention takes into account the connection between compound faults and single faults, can effectively overcome the adverse effects of the difficulty in obtaining composite fault samples with labels on diagnosis, and can directly use a single health state sample to complete the training of the model, realizing the single fault detection of rolling bearings Diagnosis and diagnosis of complex faults.

Description

technical field [0001] The invention belongs to the technical field of rolling bearing fault diagnosis, in particular to a rolling bearing composite fault diagnosis method embedded in a fault semantic space. Background technique [0002] Rolling bearings are one of the core components of many mechanical equipment. Once a fault occurs, it will greatly affect the operation of the equipment, cause economic losses, and even endanger personal safety in severe cases. Therefore, its healthy operation is very important. [0003] Compound faults have complex characteristics and are one of the difficulties in intelligent fault diagnosis. However, in engineering practice, the composite fault monitoring samples during the operation of mechanical equipment are far less than the single fault monitoring data samples, and there are even composite fault types that have not occurred during the early operation of the equipment. Under the influence of the above factors, it is difficult to obta...

AI Technical Summary

Problems solved by technology

There are significant limitations in the existing intelligent diagnosis methods for compound faults: ① directly define compound faults as a new type of health state, ignoring the connection between compound faults and single faults; ② require a sufficient number of samples in the training set Label Composite Failure Sample

Therefore, affected by the missing data set of labeled compound fault samples, the existing intelligent diagnosis methods for bearing compound fault samples identify the compound fault samples as a single fault or normal state, resulting in missed or misdiagnosed faults, resulting in low overall accuracy of the model

Images