Bridge cable wire breakage signal identification method and system based on long short-term memory network

Abstract

The invention belongs to the technical field of bridge cable state monitoring, and particularly relates to a bridge cable broken wire signal identification method and system based on a long short-term memory network, signal feature extraction is carried out from multiple dimensions of time domain, frequency domain, time-frequency analysis and the like, and feature parameters with relatively high classification capability are selected; a comprehensive feature vector representing the acoustic emission signal is constructed; a broken wire signal identification model is constructed based on LSTM, and good performance is shown on a test set; compared with a traditional machine learning algorithm model, the constructed broken wire signal identification model can accurately identify most broken wire and non-broken wire signals, and shows good identification capability for broken wire signals.

Description

Technical field [0001] The present application belongs to the field of bridge cable condition monitoring technology, specifically involving a bridge cable wire breaking signal recognition method and system based on long short-term memory network. Background [0002] The statements in this section merely provide background technical information related to this application and do not necessarily constitute prior art. [0003] With the vigorous development of bridge construction, bridge operation and maintenance safety is particularly important. Affected by the long-term service of the bridge and natural factors, the cable as a stress component of large bridges such as cable-stayed bridges, its safety and durability will be reduced. Therefore, the realization of health monitoring of bridge cables is the key to ensuring the normal operation of bridges. Acoustic emission is a dynamic non-destructive testing method, which is increasingly widely used in local monitoring such as bridge c...

AI Technical Summary

Problems solved by technology

[0005] (1) Most researchers have few analysis parameters for acoustic emission signals, and generally only analyze common acoustic emission parameters, and use parameter changes to represent the degree of damage or use one or two indicators that represent obvious changes in damage to carry out early warning of damage , some studies only distinguish damage based on statistical analysis of parameters, and the number of features extracted from acoustic emission signals is limited, which will have a certain impact on the final identification

[0006] (2) The current AE monitoring research seldom uses machine learning algorithms to identify broken wire signals; even if machine learning algorithms are used for modeling and analysis, most of them use general classification algorithms, such as clustering algorithms and neural networks Wait

Images