Fatigue factor recessive anomaly detection and fault diagnosis method based on LSTM

Abstract

The invention discloses a fatigue factor recessive anomaly detection and fault diagnosis method based on LSTM, and the method comprises the following steps: S1, collecting the time sequence data of target equipment of a related diagnosis object, and carrying out the empirical mode decomposition of the data; S2, constructing an equipment vibration signal prediction model based on LSTM by using thenormal data; S3, classifying the collected abnormal data, and constructing a fault time series data classification model based on LSTM; S4, taking the obtained mean square error MSE of the LSTM-basedvibration signal prediction model as an initial fatigue factor threshold; S5, predicting the equipment production data by using an LSTM-based vibration signal prediction model, calculating a mean square error between a predicted value and an actual value, and comparing the mean square error with an initial fatigue factor threshold to detect an abnormal signal; S6, classifying the abnormal signalsthrough a fault time series data classification model to obtain a fault diagnosis result.

Description

technical field [0001] The invention relates to the field of equipment fault diagnosis, in particular to an LSTM-based hidden abnormality detection and fault diagnosis method for fatigue factors. Background technique [0002] With the "nuclear fusion" explosion of Internet of Things, 5G, artificial intelligence, cloud computing and other technologies, the "re-industrialization" strategies formulated by major industrial countries around smart manufacturing are also rampant. In the 2019 government work report, China put forward the concept of "smart +" for the first time, and identified smart manufacturing as an important development direction of new kinetic energy for national economic development. [0003] The current fault diagnosis methods are mainly classified into two categories: methods based on signal processing and methods based on machine learning. Common methods based on signal processing include Spectral Kurtosis, Sparse Decomposition Analysis, Time-frequency Anal...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problems caused by equipment data imbalance, high noise, timing and other characteristics in the field of intelligent manufacturing, the present invention provides a method for detecting and diagnosing hidden anomalies of fatigue factors based on LSTM

Images