基于车桥耦合分析与神经网络的桥梁损伤识别

Bridge Damage Identification Based on Vehicle Bridge Coupling Analysis and Neural Network

开发一种基于车桥耦合振动分析与神经网络技术的桥梁结构损伤识别方法。以车辆行驶引起的振动响应为输入,以结构损伤状态为输出,建立能够实现桥梁损伤识别的神经网络。采用数值仿真模拟,首先设定桥梁可能发生的损伤模式,利用自主研发的车桥耦合分析程序,计算桥梁在假定损伤模式下的车致动力响应;反过来将得到的振动响应作为输入训练数据,假定损伤状态作为输出目标,建立神经网络从而实现桥梁的结构损伤识别。使用简支梁桥和平面车辆模型阐述所提出方法的基本理论并进行可行性评估。

The paper developes a bridge structure damage identification method based on vehicle bridge coupling vibration analysis and neural network technology. With the vibration response caused by vehicle driving as the input and the structural damage state as the output,a neural network that can realize bridge damage identification is established. Using numerical simulation,the paper firstly sets the possible damage mode of the bridge and calculates the vehicle induced dynamic response of the bridge under the assumed damage mode using the self-developed vehicle bridge coupling analysis program. In turn,the obtained vibration response is taken as the input training data,and the damage state is assumed as the output target. Then,a neural network is established to realize the structural damage identification of the bridge. The paper uses simple supported beam bridge and a plane vehicle model to illustrate the basic theory of the proposed method and evaluate its feasibility.