Axis track identification method based on cloud computing and LSTM

Abstract

The invention relates to an axis track identification method based on cloud computing and LSTM. The method comprises the following steps: (1) collecting vibration data; (2) preprocessing the original data, and extracting and decomposing effective IMF components; (3) inputting the IMF component of the training data into a reverse cloud generator for calculation, obtaining three digital features of the training data and forming a feature vector; (4) inputting the feature vector into a neural network LSTM model for operation, realizing identification and classification, and obtaining a training model; (5) inputting the extracted IMF components of the test data or other vibration fault data into a reverse cloud generator for calculation, and obtaining feature vectors of the IMF components; and (6) inputting the feature vector in the step (5) into a neural network LSTM model for operation, and finally obtaining an identification and classification result. According to the invention, the type of the axis track can be rapidly identified, the axis track identification accuracy is improved, an effective basis is provided for steam turbine set vibration fault diagnosis, and therefore potential safety hazards caused by vibration faults are reduced.

Description

Technical field [0001] The present invention relates to the field of thermodynamic power generation techniques, and more particularly to an axial trajectory recognition method based on cloud computing and LSTM. Background technique [0002] At present, the depth peak of the industry in the industry has become a normal, and the vibration problem is one of the key and difficulties of the long-term safe operation of the relational unit, and the axial trajectory graphics are widely used in steam turbine vibration fault diagnosis. Different axial trajectories correspond to different The vibration fault type can effectively improve the accuracy of fault diagnosis. [0003] There is a problem with the amount of fault data in the existing data recognition method, the operation cycle is long, and the time series is short, resulting in low accuracy of the axial track recognition. Therefore, there is a need for a method of accurately identifying a classification axis trajectory to improve v...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method for identifying the axis trajectory based on cloud computing and LSTM. Through the identification of the axis trajectory, the accuracy of the axis trajectory identification can be enhanced, the efficiency of fault diagnosis can be further improved, and the problem of large amount of fault data in data identification can be solved. Problems such as long operation cycle and short failure time selection sequence

Images