用人工神经网络方法估计桥梁在温度作用下的挠度行为

Prediction of Bridge Structural Performance under Temperature by Artificial Neural Network

桥梁结构的挠度变化除了与车辆、人群等荷载有关之外,还与环境的作用密切相关,在大跨径桥梁结构安全监测系统中进行结构安全状态评估时需要知道桥梁结构在环境作用下的行为,以便分离环境与例如汽车、人群等荷载的不同影响。挠度变化是桥梁结构安全监测系统中反应桥梁安全状态最直观的参数之一,荷载与环境的作用、结构材料的变异都可以通过挠度的变化表现出来。但是通常所测得的挠度是桥梁在环境温度、车辆荷载等综合作用下的总响应,而不同荷载条件作用可以使得结构产生同样的挠度,由于不同的荷载条件可以引起同样的挠度但是产生不同的应力,因此将挠度分类是十分必要的。本文尝试用神经网络方法通过实测值来模拟温度与挠度之间的非线性关系,并用它来预测桥梁由温度所产生的挠度变化,从而可以将这温度产生的挠度值从实时的总挠度中分离出来,以便进行其他部分挠度评估的分析。

Bridge deflection is the most intuitionistie parameter in the bridge monitoring system, which reflects the effect of vehicle load and environment around the bridge. While analyzing field measured deflection of a bridge, it is necessary to separate the integrate response of the deflection caused by different effects. The authors try to use artificial neural network method to simulate the nonlinear relations between temperature and deflection of bridges, and predict the deflection of bridge under certain distribution of temperature. The deflection can be separated from the whole response.