Fault diagnosis of elevator worm gear box bearing based on feature fusion

In order to solve the problem of fault diagnosis of elevator worm gear box bearing, the method based on continuous wavelet transform and long short-term memory neural network is adopted. The collected data is first filtered by an infinite impulse response filter, and then the scale maps of the current signal and the vibration signal are obtained by continuous wavelet transform and spliced at the same time, and finally they are put into the long short-term memory neural network for analysis and training. The experimental results indicate that the accuracy of the proposed approach for the fault diagnosis of elevator worm box bearings can reach 99.85%.