Mechanical equipment fault diagnosis method based on multiple measurement vectors of wireless sensor network

Abstract

The invention relates to a mechanical equipment fault diagnosis method based on multiple measurement vectors of a wireless sensor network, and belongs to the technical field of deep learning. The method comprises the following steps: S1, compression acquisition of multiple measurement vectors; s2, data processing; s3, training of a convolutional neural network model; and s4, identifying of fault types. According to the method, the limitation of the Nyquist sampling law is broken through, and the data acquisition amount is reduced, so that the load of the wireless sensor network is reduced. Compared with a traditional mode of manually extracting features for diagnosis, the method is more convenient and efficient, and the requirements for manpower and professional knowledge are reduced. Datacompression acquisition is carried out by adopting multiple measurement vectors MMV, so that the acquired compressed data volume only contains more fault source information, and the identification rate is higher than that of fault diagnosis of an SMV model.

Description

technical field [0001] The invention belongs to the technical field of deep learning, and relates to a mechanical equipment fault diagnosis method based on wireless sensor network multi-measurement vectors. Background technique [0002] Large and complex mechanical equipment, such as: natural gas pipelines, bridges, large motors and machine tools, etc., once these mechanical equipment fail and are not eliminated or repaired in time, their failure can lead to huge economic losses and extremely serious consequences. It is very important to effectively evaluate and predict the health status of these mechanical equipment, as well as to carry out timely fault diagnosis and identification. [0003] The patent number is CN110991295, which discloses an adaptive fault diagnosis method based on one-dimensional convolutional neural network. The one-dimensional time series signal is input into the one-dimensional convolutional neural network, and the diagnosis result can be obtained. Co...

AI Technical Summary

Problems solved by technology

This method is also not suitable for large-scale mechanical equipment, because the compressed collection method of this method is Single Measurement Vector (Single Measurement Vector, SMV), and its data volume only contains part of the fault source information, resulting in a low fault identification rate

Images