Studies on Damage Distinguishing of Underground Tunnel Structure Based on BP Neural Network

Abstract

With the expansion and development of scale of construction on metro engineering, the damage diagnosis and the safety evaluation on underground engineering structure have become vital problems to be solved. This paper raised an idea to distinguish underground engineering structure based on BP neural network: define change rate of curvature of structure, and recognize it as the input scalar of BP neural network, using a reducing unit elastic modulus method to simulate damage location and damage degree, through various set of underground structure extent of damage, recognize the first four order curvature structure change rate as input of BP neural network. The results show that the method using BP neural network can identify the damage degree of underground engineering structure accurately and can solve the damage identification problem of underground engineering structure conveniently and effectively.