Method and device for fault prediction of industrial equipment based on lstm cyclic neural network

Abstract

The invention discloses a fault prediction method and device for industrial equipment based on an LSTM circulating neural network, wherein the method comprises the following steps of: acquiring a state monitoring data set of a plurality of sensors at the periphery of target equipment, wherein the state monitoring data set comprises monitoring data from 0 moment to a current moment; selecting a prediction characteristics containing preset fault information from the state monitoring data set by utilizing a characteristic selection standard, wherein the characteristic selection standard comprisesa correlation index and a monotonicity index; performing characteristic conversion on the prediction characteristics to obtain a prediction characteristic vector; and performing single-step fault prediction, long-term fault prediction and residual life prediction on the target equipment according to the prediction characteristic vector and a fault prediction network model. The method can effectively avoid insufficient prediction precision caused by unreasonable preset fault threshold, can give a confidence interval under the occasion of single-step performance prediction, and can achieve long-term prediction of performance and residual service life of the equipment.

Description

technical field [0001] The invention relates to the technical field of data-driven fault prediction, in particular to a method and device for fault prediction of industrial equipment based on LSTM (Long Short-Term Memory, long-short-term memory network) cyclic neural network. Background technique [0002] Related technologies, data-driven equipment failure prediction methods are mainly based on statistical analysis, Bayesian network, SVM (Support Vector Machine, support vector machine), HMM (Hidden Markov Model, hidden Markov model) and NN (Neural Network, neural network) and so on. Although the above methods have achieved good results in specific tasks and scenarios, there are problems such as low prediction accuracy, insufficient promotion ability, and difficulty or inability to make long-term predictions. They should be applied to follow-up work such as operation and maintenance strategy optimization. There are difficulties. [0003] The main forms of fault prediction a...

AI Technical Summary

Problems solved by technology

Existing RNN-based fault prediction methods are generally implemented in two ways: one takes the features at time t, t-T, t-2T... However, with the increase of T, the prediction accuracy will be significantly reduced, and when T is small, it will lose the ability to guide maintenance decisions, so it is difficult to apply in practice; the other uses RNN as a feature extraction model, based on the features input by RNN, Generally, the remaining life is calculated through the preset fault threshold and exponential model, so the advantages and characteristics of RNN compared with other neural networks have not been fully utilized

Images