基干路表弯沉盆BP人工神经网络反演沥青面层弹性模量探讨.docVIP

基于路表弯沉盆的BP人工神经网络反演沥青面层弹性模量研究 　　摘 要：基于层状弹性体系理论，建立BP人工神经网络反演沥青路面沥青面层弹性模量预测模型，利用BP人工神经网络预测沥青路面沥青面层弹性模量。理论弯沉盆和实测弯沉盆反演沥青面层弹性模量的结果表明，建立的BP人工神经网络反演沥青路面沥青面层弹性模量模型具有良好的预测精度和可靠性，为评价沥青路面的沥青面层性能状况提供了参考 关键词：BP人工神经网络 路表弯沉盆 弹性模量反演 沥青面层 沥青路面 中图分类号：U41 文献标识码：A 文章编号：1674-098X（2015）10（b）-0091-04 Research on elastic modulus backcalculation of asphalt course using BP artificial neural network based on surface deflection basin of pavement Yang Guoliang Zhong Wen Huang Xiaoyun Liang Simin He Huihui Chen Jiaju （School of Civil Engineering of Guangzhou University，Guangzhou Guangdong，510006，China） Abstract：Based on layered elastic theory，the elastic modulus of asphalt course in asphalt pavement was predicted using BP artificial neural network.According to the types of pavement structure in common use，the database of surface deflections with their corresponding structural parameters of asphalt course based on layered elastic theory was established.The elastic modulus backcalculation model of asphalt course in asphalt pavement was developed using BP artificial neural network to predict.The predictive results of asphalt course elastic modulus backcalculation using theoretical deflection basin and measured deflection basin indicate that the elastic modulus backcalculation model of asphalt course in asphalt pavement is of good predictive accuracy and reliability.It would provide the references with the elastic modulus backcalculation model of asphalt course to accurately and quickly estimate the conditions of asphalt course in asphalt pavement. Key Words：BP artificial neural network；Ssurface deflection basin of pavement；Elastic modulus backcalculation；Asphalt course；Asphalt pavement 表征沥青面层性能的其中一个最为重要的指标就是其弹性模量。众多的国内外研究[1-2]表明，利用落锤式弯沉仪采集的路表弯沉盆反算沥青面层弹性模量是可行的。因此，文章基于层状线弹性体系理论，并结合BP人工神经网络具有并行处理、很强的高度非线性映射能力等特点[3-4]，进行了采用BP人工神经网络由路表弯沉盆反演沥青面层弹性模量的探讨 1 沥青路面结构力学分析模型 假设路面沿深度方向分成为n个水平层，层间满足连续条件，路面表面作用有圆形均布轴对称垂直荷载p，第i层厚度、弹性模量和泊松比分别定义为hi，Ei，μi（i=1，2，…，n-1），最下层为土基弹性半空间体，其弹性模量和泊松比分别