基于复合分形的路面抗滑随机森林评估模型

A Random Forest Evaluation Model for Pavement Skid Resistance Based on Comprehensive Fractal

路面抗滑性能直接影响着道路交通安全,而基于路表纹理特征的路面抗滑性能评估方法目前存在着可解释性差、准确度不高的问题.该研究使用精度为0.05 mm的便携式三维激光表面分析仪采集了185组路面纹理数据,通过动态摩擦因数测试仪获得了相应路段0~80 km/h速度范围内的路面摩擦数据,构建了综合表征路面纹理空间、横剖、深度方向复杂度的复合分形维数指标,建立了10 km/h和70 km/h速度下的路面抗滑性能随机森林评估模型.研究结果表明:复合分形维数具备独立描述纹理复杂程度的能力,但与路面动态摩擦因数之间不存在线性关系;复合分形维数对70 km/h速度下动态摩擦因数预估的准确度为0.78,可用于评价轮胎橡胶快速滑动状态下的路面抗滑性能;复合分形指标中的空间、横剖、表层、浅层、深层剖面分形特征共同影响着路面抗滑性能,在进行路面纹理形貌评价时,应从多种空间视角下进行纹理特征综合分析.

The pavement anti-skid performance directly affects road traffic safety,and the evaluation methods based on pavement texture features currently have problems of poor interpretability and low accuracy.Herein,185 sets of pavement texture data were collected by the portable 3D laser surface analyzer with an accuracy of 0.05 mm.The pavement friction data in the speed range of 0~80 km/h of the corresponding road section were obtained with the dynamic friction coefficient tester.The comprehensive fractal dimension index representing the complexity of the pavement texture space,the cross section,and the depth direction was constructed,and the random forest evaluation model for pavement skid resistance performances at speeds of 10 km/h and 70 km/h.The results show that,the comprehensive fractal dimension has the ability to describe the complexity of texture independently,but there is no linear relationship between it and the pavement dynamic friction coefficient;the prediction accuracy of comprehensive fractal dimensions for dynamic friction coefficients at the 70 km/h speed is 0.78,which can be used to evaluate the skid resistance of pavement under the condition of rapid sliding of tire rubber;the spatial,cross-sectional,surface,shallow,and deep profile fractal features in comprehensive fractal indicators jointly affect the pavement anti-skid performances.In the evaluation of pavement texture morphology,comprehensive analysis of texture features should be conducted from multiple spatial perspectives.