基于IGWO-XGBoost融合模型的沥青路面抗滑性能评估

为研究沥青路面抗滑性能影响因素,精确预测路面抗滑性能,本文使用Gocator 3110三维智能传感器采集沥青混合料试件表面纹理并使用摆式摩擦仪测试试件表面摩擦系数.针对三维纹理点云数据中的异常数据,提出基于径向基函数(RBF)的邻域插值算法进行数据质量提升.根据修复后的三维纹理点云数据计算出具有代表性的10类宏观纹理特征参数,并采用Pearson系数相关性分析法去除冗余因子,改进模型的输入特征,并构建基于改进灰狼优化算法(IGWO)与XGBoost融合的沥青路面抗滑性能预测模型,预测沥青路面的摩擦系数.结果表明,提出模型的预测精度优于多元线性回归模型、支持向量机回归模型以及基于网格化搜索的XGBoost模型.

Evaluation of Skid Resistance of Asphalt Pavement Based on IGWO-XGBoost Fusion Model

To study the influencing factors on the skid resistance of asphalt pavements and accurately predict the skid resistance of the pavements, this study resorts to the Gocator 3110 three-dimensional (3D) intelligent sensor to obtain the surface textures of asphalt mixture specimens and employs the pendulum friction tester to measure the surface friction coefficient of the specimens. Regarding the abnormal data in the 3D point cloud data of the textures, the study proposes a radial basis function (RBF)-based neighborhood interpolation algorithm to improve data quality. Then, the feature parameters of 10 typical macro-textures are calculated with the reconstructed 3D point cloud data of the textures, and the redundancy factor is removed by Pearson coefficient correlation analysis to improve the input features of the model. A prediction model for the skid resistance of asphalt pavements integrating an improved gray wolf optimization (IGWO) algorithm and XGBoost is developed to predict the friction coefficient of asphalt pavements. The results show that the prediction accuracy of the proposed model is better than that of the multiple linear regression model, the support vector machine regression model, and the XGBoost model based on grid search.