基于遗传算法的SMA路面抗滑性能预测模型  

Prediction Model for Skid-resistance of Stone Mastic Asphalt Pavement Based on Genetic Algorithm

作　　者：刘亚敏[1] 韩森[1] 徐鸥明[2] 

机构地区：[1]长安大学特殊地区公路工程教育部重点实验室,陕西西安710064 [2]长安大学材料科学与工程学院,陕西西安710061

出　　处：《应用基础与工程科学学报》2013年第5期890-898,共9页Journal of Basic Science and Engineering

基　　金：陕西省交通科技项目(KY08-01);中央高校基本科研业务费专项资金项目(2013G1211007)

摘　　要：为了量化各因素与SMA(Stone Mastic Asphalt)路面抗滑性能的相关性,即建立抗滑性能预测模型,首先利用参数辨识法,确定预测模型的表达式形式和变量,然后根据遗传算法的步骤,依次进行参数的编码和群体设定,适应度函数的确定,交叉概率、变异概率和算法终止准则的确定,各变量初值的确定等工作,最终建立了各因素,包括最大粒径通过率、最大粒径次一级筛孔通过率、关键筛孔通过率、关键筛孔次一级筛孔通过率、沥青用量等与SMA路面抗滑性能之间的预测模型.精度检验结果表明,实测值与计算值无显著差异,模型较为合理,可以用于指导SMA配合比设计.In order to obtain a quantized relationship between several factors and the skid-resistance of SMA （Stone Mastic Asphalt）, which means building a prediction model, firstly according to parameter identification method, the formula form and variables for the prediction model were selected. Secondly based on genetic algorithm, factors＇ coding and population＇s setting, selection of the fitness function, determination of the crossover probability, mutation probability and the algorithm＇s stop criterion, selection of the factors＇ initial value were carried out in turn. Lastly the model between skid-resistance and factors was obtained. The accuracy test results indicated that there was not remarkable difference between test value and calculated value, and the model was logical and useful for the mixture design of Stone Mastic Asphalt.

关 键 词：道路工程 抗滑性能 预测模型 遗传算法 SMA 

分 类 号：U414.3[交通运输工程—道路与铁道工程]