沥青路面智能压实多元回归模型建立与验证

Establishment and Verification of Multiple Regression Model of Intelligent Compaction for Asphalt Pavement

量化分析智能压实过程中碾压参数对沥青路面压实程度的影响,能够科学地指导沥青路面压实工作。依托两种不同的沥青路面(AC-20和SMA-13)开展智能压实试验,采集了振动压实值(VCV)与智能压实参数等相关数据;基于多元回归模型原理与方法,分别建立了两种沥青路面的智能压实多元线性回归模型和多元非线性回归模型并进行了验证。结果表明:多元线性回归模型和多元非线性回归模型均能较好地预测两种沥青路面在智能压实过程中的VCV值变化情况;与多元线性回归模型相比,多元非线性回归模型能够更好地预测沥青路面的智能压实状态。

The quantitative analysis of the influence of rolling parameters on the compaction degree of asphalt pavement in the process of intelligent compaction can scientifically guide the compaction of asphalt pavement.The intelligent compaction test is carried out on two different asphalt pavements(AC-20 and SMA-13).The vibration compaction value(VCV)and related intelligent compaction parameters are collected.Based on the principle and method of multiple regression model,multivariate linear model and multivariate nonlinear model of intelligent compaction for the two kinds of asphalt pavement are established and verified.The results show that both multivariate linear model and multivariate nonlinear model can predict the VCV changes of the two asphalt pavements during the process of intelligent compaction.Compared with multivariate linear model,multivariate nonlinear model can better predict the intelligent compaction state of asphalt pavement.