Bridge vibration modal visual damage identification method based on machine learning

Abstract

The invention discloses a bridge vibration modal visual damage identification method based on machine learning, and the method comprises the following steps: S1, collecting the digital video information of bridge vibration, and carrying out the amplification processing of the digital video information; S2, acquiring parameters of the digital video image information, wherein the parameters comprisefrequency, phase and amplitude of the video image information; S3, constructing a visual modal image of the bridge vibration according to the phase and amplitude of the video image information; and S4, extracting multi-level feature information in the bridge visual modal image by using the convolutional neural network CNN, inputting the extracted feature information into a generalized regressionneural network GRNN, and carrying out identification estimation on the damage property of the bridge structure. According to the bridge vibration mode visual damage identification method based on machine learning, the structure of the bridge is detected through a small number of simple devices, the structural damage of the bridge can be accurately and comprehensively identified, and the method isscientific and efficient.

Description

technical field [0001] The invention relates to a method for structural engineering and safety monitoring, in particular to a method for visual damage identification of bridge vibration modes based on machine learning. Background technique [0002] With the gradual increase in the number of bridges built in our country, it is undeniable that our country has entered the ranks of bridge powers. However, due to our special national conditions and frequent domestic overloading, some bridges are aging ahead of time, bridge structural materials are aging, loads, corrosion, Fatigue effects and bridge management deficiencies, coupled with various unforeseen natural disasters, have made the fatigue damage of bridge structures more and more serious, and gradually manifested in many problems such as reduced performance and insufficient bearing capacity, which have become potential threats to modern traffic, affecting The health and normal use of the bridge. In this regard, the damage ...

AI Technical Summary

Problems solved by technology

And most of the damage methods are to test and collect data locally on the key points of the structure. The method mostly uses contact measurement methods, such as installing various sensors on the bridge structure. Not only the test range is relatively local,

Images