基于BP神经网络桥梁的损伤识别方法

Damage Identification Method of Bridge Based on BP Neural Network

在桥梁损伤识别研究中,常常不能对桥梁损伤单元进行准确定位。为使用更简便的方法对桥梁进行精确的损伤定位,提出借助BP神经网络的自我学习功能的方法,将桥梁某点的挠度影响线数据、桥梁损伤单元号及损伤单元剩余刚度系数,作为BP神经网络的运行参数,进而训练出一个可以通过输入桥梁挠度影响线数据来定性及定量识别桥梁单点损伤的方法。依托工程实例对所提损伤识别方法的可行性进行了验证。

In bridge damage identification studies,accurate localization of bridge damage units is often not possible.In order to further use a simpler method for accurate damage localization,a method was proposed to qualitatively and quantitatively identify a single point of bridge damage in the paper,by inputting bridge influence line data,bridge damage unit number and residual stiffness coefficient of the damage unit as the operating parameters of the BP neural network with the help of the self-learning function of the BP neural network.The feasibility of the proposed damage identification method is verified by relying on engineering examples.