Motor fault detection method based on group type sparse self-coding and swarm intelligence

Abstract

The invention discloses a motor fault detection method based on group-type sparse self-coding and group intelligence, and the method comprises the steps: (1) controlling a magnetic powder brake through the output current of a tension controller to achieve the adjustment of the load torque of a motor for a fault signal of a DC motor, and collecting the fault signal of the motor; (2) establishing a group type sparse auto-encoder implementation framework, solving by using an M-M method, and performing feature extraction on original input motor fault data by using the sparse feature extraction network; (3) improving a swarm intelligence algorithm, and proposing an SS-PSO algorithm; (4) establishing an SS-PSO-ANN depth classifier model; and (5) for high-quality sparse features extracted based on the group type sparse auto-encoder implementation framework network, performing effective diagnosis of motor faults by using an SS-PSO-ANN classifier. According to common faults of turn-to-turn short circuit and rotor demagnetization of the direct current motor, autonomous fault diagnosis of the motor is realized, so that the situation that the operation efficiency of the motor is influenced by the faults and even accidents are caused by damage of the motor is avoided.

Description

technical field [0001] The invention relates to the technical field of motor fault diagnosis, in particular to a motor fault detection method based on group sparse self-encoding and group intelligence. Background technique [0002] Electric motors have become the most widely used basic power equipment in industrial production. Motor failure not only affects production, but can also cause major safety incidents. Therefore, it is particularly important to diagnose and eliminate faults in time, prevent accidents, and ensure safe, reliable and efficient operation of motors. [0003] Traditional fault diagnosis methods include methods based on wavelet packet analysis, eigenvector methods or methods based on Hilbert-Huang transform. However, due to the complex fault mechanism of the motor, the fault characteristic signal is not obvious, and the fault signal is nonlinear and non-stationary, which makes the above traditional feature extraction methods have some defects. For exampl...

AI Technical Summary

Problems solved by technology

However, due to the complex fault mechanism of the motor, the fault characteristic signal is not obvious, and the fault signal is nonlinear and non-stationary, so there are some defects in the above traditional feature extraction methods. Based on signal processing methods such as Hilbert-Huang transform or wavelet transform, the generalization ability of fault feature extraction is subject to certain constraints, which cannot meet the current requirements of motor fault monitoring.

Images