基于NAR动态神经网络的桥梁SHM应变预测

Bridge SHM Strain Prediction Based on NAR Dynamic Neural Network

基于NAR动态神经网络对桥梁SHM应变进行了分析预测。分别建立了10期、25期、50期的NAR动态神经网络预测模型和ARIMA预测模型,并对比了2个模型预测的稳定性和准确性,通过上海市某座斜拉桥的SHM实测应变数据进行验证。模型结果表明,NAR动态神经网络预测模型相比于ARIMA预测模型具有更高的准确性,工程应用价值更高。

Based on NAR dynamic neural network,SHM strain of bridge is analyzed and predicted.The NAR dynamic neural network prediction model and ARIMA prediction model of 10,25 and 50 stages are established respectively,and the stability and accuracy of the two models are compared,which are verified by the SHM strain data of a cable-stayed bridge in Shanghai.The results show that NAR dynamic neural network prediction model has higher accuracy and engineering application value than ARIMA prediction model.