Evolution Trend Prediction Technique For Digital Twin Model Of Section-Level Asphalt Pavement

At present,the road maintenance mileage in China has reached more than 5 million kilometers.Promoting the digital and intelligent maintenance of asphalt pavement is an effective way to improve the efficiency of road maintenance management.Using digital twin technology to build a virtual model with the same value as the physical entity can realize the simulation analysis,prediction and optimization of the physical entity.At present,the provincial highway management organizations have established the digital twin of traffic infrastructure,realizing the mapping from physical entity to digital twin through digital twin technology.However,in the maintenance work,how to realize the digital maintenance of diseases in asphalt pavement according to the disease data in the digital twin is still being explored.During the application of asphalt pavement digital twins to maintenance,in addition to realizing the visual display of diseases,the storage of disease information and the calculation of pavement damage technical condition index,the evolution trend prediction of digital twins should be completed according to the data in asphalt pavement digital twins,so as to provide support for disease maintenance decision-making.Based on this,this paper carries out the research on the prediction technology of the evolution trend of digital twins of section level asphalt pavement.Firstly,the related concepts of asphalt pavement digital Twin are defined,the framework of asphalt pavement digital twin is proposed,and the construction basis of asphalt pavement digital twin is analyzed;Then,the influencing factors of asphalt pavement damage are investigated,and the ways to obtain the representative values of the influencing factors are described.The method to divide the asphalt pavement digital twin sections according to the inherent attributes of the influencing factors is proposed,and the construction method of the section level asphalt pavement digital twin is proposed.Secondly,the technical framework of disease evolution trend prediction of asphalt pavement digital twins is proposed;Then,facing the digital twin,the disease is re divided into structural disease and apparent disease.According to the disease description-disease category-disease cause,the disease cause chain tree is constructed,and the intelligent matching algorithm is used to automatically identify the disease category and cause.The disease prediction model is determined according to the disease type,development law and corresponding cause,and the transverse cracks and ruts are used to represent the structural disease and apparent disease,The above methods are illustrated by examples.Finally,according to the actual disease situation of Expressway in Jiangsu Province,taking the rutting disease as an example,the application of the segment level digital twin evolution trend prediction technology is analyzed,and compared with the traditional rutting prediction model modeling methods,the characteristics of the three prediction methods are analyzed.Firstly,the traditional finite element model for rutting disease prediction and the establishment method of rutting prediction model based on mechanical empirical model are described;Then,after collecting the rutting disease information through the automatic detection equipment,the rutting disease detection data is mapped to the digital twin of the road section level asphalt pavement,and the convolution neural network algorithm is used as the core algorithm of the rutting type identification module to complete the rutting type identification and cause analysis,so as to quickly select the appropriate rutting depth prediction model in the disease prediction model base,The application analysis of digital twin evolution trend prediction of section level asphalt pavement is completed,and the applicability of the three methods to rutting depth prediction is compared.