Elevator fault early warning method based on bidirectional gated recurrent neural network

Abstract

The invention discloses an elevator fault pre-warning method based on a bidirectional gating circulating neural network. The elevator fault pre-warning method comprises the following steps: (1) sampling elevator vibration waveform data; (2) converting elevator vibration waveform samples into a sequence form; (3) classifying serialized samples into a training set and a testing set; (4) constructinga bidirectional gating circulating neural network framework; (5) training the bidirectional gating circulating neural network framework; (6) carrying out prediction testing by virtue of the testing set, so as to obtain a bidirectional gating circulating neural network prediction model; and (7) carrying out elevator fault prediction classification by virtue of the bidirectional gating circulatingneural network prediction model, so as to obtain a predicted elevator waveform result. The invention further discloses an elevator fault diagnosis method based on the bidirectional gating circulatingneural network. When a judgment result shows that an elevator is in a fault state, an alarm is given out. The two methods disclosed by the invention are extremely high in accuracy rate and instantaneity on the diagnosis and prediction of elevator conditions.

Description

technical field [0001] The invention belongs to the technical field of elevator fault detection and early warning, relates to an elevator fault prediction method based on a bidirectional gating cyclic neural network, and also relates to an elevator fault diagnosis method based on a bidirectional gating cyclic neural network. Background technique [0002] In recent years, with the continuous development of social economy, urban high-rise buildings have increased rapidly, elevators have been widely used, and the number of elevators has continued to increase. In the future, the elevator update market and after-sales service market will have huge space. However, in recent years, elevator failures have occurred frequently. On the one hand, the high incidence and severity of accidents often cause considerable casualties and large economic losses; Conditions and maintenance records are kept unevenly, making it difficult to verify a lot of real data, and burying a lot of safety haza...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide an elevator fault prediction method based on a two-way gated cyclic neural network, which solves the problem that the prior art cannot predict the occurrence of faults in advance, and can only be rescued passively. Questions that reflect the health of the elevator