Construction of Data-Driven Explicit Rutting Evolution Model for Asphalt Pavement

Source abstract

Rutting is a non-stationary complex process, and its research has been the focus and difficulty in the world and industry for a long time. Effectively revealing the evolution law of asphalt rutting will play a positive role in optimizing road design, preventive maintenance and resource saving, which will help promote the development of transportation infrastructure to a more environmentally friendly and low-carbon direction. How to capture the temporal and spatial evolution law of rutting and express it through an explicit model will greatly improve the research level of permanent damage of asphalt pavement, which is the research difficulty of the industry and also the focus of this paper. In this paper, RIOHTrack was used to measure the database of more than 100 million loads of full-scale ring road. Time series method was used to overlay the original Kou&Cao model framework proposed by our team, and an explicit model framework was proposed here, which can well present the linear and nonlinear process of rutting evolution in time and space. The model framework perfectly fits the process of rutting, the fitting accuracy reaches 0.993 when tested on seven kinds of pavement structure data. The rolling prediction accuracy reaches 0.972. The proposed model framework effectively improves the interpretability of the asphalt rutting evolution model, greatly improves the accuracy of the asphalt rutting evolution model, and has excellent generalization ability, which is closer to revealing the real situation of the temporal and spatial evolution of asphalt pavement, and plays an important role in the study of the long life of asphalt pavement.