Small and medium-sized bridge lightweight health monitoring method and system based on acoustic emission and 1D CNNs

Abstract

The invention discloses a small and medium-sized bridge lightweight health monitoring method and system based on acoustic emission and a one-dimensional convolutional neural network (1D CNNs). The method comprises the following steps: arranging acoustic emission sensors at key vulnerable parts of a small and medium-sized bridge for real-time data acquisition; performing empirical mode decomposition on the acquired acoustic emission data to achieve the purpose of denoising; the small and medium bridges are subjected to a reduced scale model experiment, and a 1D CNNs model is trained by using acoustic emission signals of the reduced scale model experiment. And inputting signals acquired on the real bridge into the 1D CNNs model to monitor the damage stage of the structure in real time, and performing graded early warning. The invention provides the real-time lightweight health monitoring method and system for the small and medium-sized bridge by utilizing the advantages that the acoustic emission technology can monitor structural damage in real time and the 1D CNNs can simplify the calculation complexity, and provides a basis for the safety decision of a bridge management department.

Description

technical field [0001] The invention relates to the technical field of structural health monitoring, in particular to a monitoring method and system combining acoustic emission and deep learning. Background technique [0002] Small and medium-sized bridges are widely used in our daily life. Because bridges are exposed to outdoor environment every day, they will be damaged by different degrees of sudden or cumulative damage. It is necessary to carry out scientific and convenient safety assessment for small and medium-sized bridges. In order to grasp the damage status of bridges, various damage detection methods have emerged. Compared with other nondestructive testing methods, acoustic emission technology, as a widely used nondestructive testing technology, can detect dynamic defects of complex components, and can provide overall or range rapid detection and damage early warning. However, it is difficult to grasp the safety information of the structure in real time only by de...

AI Technical Summary

Problems solved by technology

Clustering analysis needs to manually create a scoring system offline and label values, which is not intelligent enough, and it is difficult to obtain clustering conclusions when the sample size is large

Emerging deep learning strategies are expected to be able to process a large number of acoustic emission signals online to achieve safety warnings, but two-dimensional neural networks have high computational complexity and require specialized hardware for training, which is not suitable for mobile devices and low power consumption or low Real-time applications on in-memory devices

Images