Industrial motor bearing fault diagnosis method based on multi-local-model decision fusion

Abstract

The invention discloses an industrial motor bearing fault diagnosis method based on multi-local-model decision fusion. The method comprises the four steps: data collection, data processing, building of multiple local diagnosis models, and decision-level fusion. In the step of data collection, two different types of sensors are employed to collect the related data of a motor bearing. In the step ofdata processing, preprocessing and feature extraction are performed on the two kinds of collected data. In the step of building of multiple local diagnosis models, local diagnosis models for the twokinds of processed data are built based on a Bi-LSTM neural network. In the step of decision-level fusion, decision-level fusion is performed on the local diagnosis results output by the two local diagnosis models based on the DSmT theory to obtain a final diagnosis result. According to the industrial motor bearing fault diagnosis method designed by the invention, a factory can be better helped tofind motor faults caused by bearing damage timely and accurately, and the influence on motor operation and factory production efficiency due to shutdown is avoided to a certain extent.

Description

technical field [0001] The invention relates to an industrial motor bearing fault diagnosis method based on multi-local model decision fusion, belonging to the field of motor fault detection. Background technique [0002] In today's society, industrial motors generally provide power for the operation of electrical appliances or various machinery. Among them, the motor bearing failure is an urgent problem to be solved in the motor prone to failure. According to statistics, 40-70% of electromechanical transmission system and motor failures are caused by rolling bearing damage, which may cause a great impact on the operation of the motor due to downtime, which in turn will affect the production efficiency of the factory and cause a certain degree of damage The economic impact, what's more, cause large-scale accidents, so that the personal safety of operators is threatened. Therefore, it is necessary to check the various unstable or fault states of the motor bearings, predict ...

AI Technical Summary

Problems solved by technology

Most traditional intelligence methods belong to shallow learning models, which are difficult to effectively learn complex representative features and nonlinear relationships

Images