Design of damage identification algorithm for mechanical structures based on convolutional neural network

Abstract

SummaryDamage identification, location, and estimation of engineering structures are very popular research topics in recent years. Structural damage detection technology has been widely used in aerospace, civil engineering, machinery, and nuclear industry. It is a multi‐disciplinary and comprehensive technology based on damage mechanism, sensor technology, signal analysis technology, computer technology, and convolutional intelligence technology. Compared with the traditional structural damage detection methods, this paper mainly studies the theory and application of structural damage detection technology based on convolutional neural network. In this paper, the damage location and damage degree of a frame structure are numerically simulated by the combined parameter method, which is suitable for structural damage identification. At the same time, the input parameters of the improved convolutional neural network are constructed by a suitable method, and the structure damage detection and identification are carried out by using the trained convolutional neural network.