Structural modal parameter identification method based on machine learning

Abstract

The invention provides a structural modal parameter identification method based on machine learning. The structural modal parameter identification method comprises the following steps: step 1, carrying out preliminary filtering and denoising processing on data collected by a vibration sensor; step 2, inputting the filtered data into a designed neural network, and designing an objective function which is used for ensuring that an output result of a third layer of the neural network has complete independence, so that a training process of the neural network is changed into a separation process of aliasing signals; step 3, extracting a result of a third layer of the neural network to obtain modal response of each order, wherein a weight among the third layer and the fourth layer of the neuralnetwork is a vibration mode coefficient of each order; and step 4, carrying out power spectrum conversion on the extracted modal response to solve the frequency, and carrying out curve fitting by utilizing a logarithm attenuation technology to obtain the damping ratio. According to the invention, a machine learning method is utilized to realize automatic processing of monitoring data, the networkautomation degree is high, and the separation speed is high.

Description

Technical field [0001] The invention relates to the technical fields of machine learning, signal processing, and civil engineering structural health monitoring, in particular to a method for identifying structural modal parameters based on machine learning. Background technique [0002] The identification of structural modal parameters (frequency, damping ratio, mode shape) is a classic inverse problem of structural dynamics. Structural modal parameters represent the dynamic characteristics of a structure and are only related to the physical parameters and mechanical models of the structure. They are of great significance in structural health detection and are the basis for structural damage identification, model update and safety assessment. In the past few decades, modal analysis has received more attention in the identification of linear systems. It decouples complex multi-degree-of-freedom systems into simple modal superpositions, so that the vibration of the structure can ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to solve the problems that the existing independence-based structural modal parameter identification algorithm has poor separation effect, easily produces modal aliasing and modal loss, and cannot meet the accuracy and efficiency requirements for accurately identifying structural modal parameters. In addition, due to the low degree of automation and high cost of manual expert intervention, a method for identifying structural modal parameters based on machine learning is proposed.

Images