Modeling Of Diseased Pavement Traffic Scenarios Based On Digital Twins

With the development of smart transportation,an increasing number of researchers are using digital twin technology to establish a mapping relationship between real and virtual traffic environments,which is used to simulate real traffic environments and improve traffic safety and efficiency.In reality,intensive road usage can lead to appear the damaged pavement,which can affect the stability and traction of vehicles.As a result,the traffic safety will be hidden danger.However,the current modeling of traffic scenarios rarely considers constructing modeling of damaged pavement.It results in discrepancies between real and virtual traffic scenes.If we can construct three-dimensional modeling of damaged pavement,it can more realistically reflect the shape and characteristics of the road,which can improve the accuracy and reliability of the simulated traffic environment.Therefore,we utilize the digital twin modeling approach and employ Structure From Motion(SFM)for three-dimensional reconstruction of damaged pavement.Additionally,we utilize real-time 3D rendering technology in Unity to construct a digital twin traffic scene featuring damaged pavement.To visualize the above content,we take Xi’an Technological University’s Weiyang Campus as the modeling object.The main contents are as follows:(1)Three-dimensional reconstruction of damaged pavement based on SFMThe mainstream methods for modeling traffic scenarios have overlooked the impact of damaged pavement on traffic safety.This paper addresses this issue by utilizing the SFM method to reconstruct damaged pavement in three-dimensions,thereby reducing the discrepancies between real and virtual traffic environments.To achieve the three-dimensional reconstruction of damaged pavement,we first use the Scale-Invariant Feature Transform(SIFT)method to perform feature matching on the images of the damaged pavement captured by a monocular camera.The Random Sample Consensus(RANSAC)algorithm is employed to eliminate erroneous match pairs.By utilizing SFM,we obtain a sparse 3D point cloud of the damaged pavement.Then,using the dense matching of the patch model,we generate a dense point cloud of the damaged pavement.Finally,we reconstruct the three-dimensional model of the damaged pavement by employing the poisson surface reconstruction method.(2)Unity3D Campus traffic scene physical model construction of damaged pavementTo address the problem for high-fidelity replication of the physical world in virtual traffic environments,this paper utilizes real-time 3D rendering technology in Unity3 D to construct a physically accurate model of a campus traffic scene with the presence of damaged pavement,which incorporate the four elements of ―people,vehicle,road,and environment‖.Firstly,the3 D modeling of the damaged pavement,buildings,and vegetation in Xi’an Technological University’s Weiyang Campus is conducted.Secondly,writing scripts to control dynamic weather changes in the urban setting,the artificial intelligence vehicle,and pedestrian movement control and obstacle avoidance function.Finally,optimization techniques are applied to all models to achieve the construction of a physically accurate model of a campus traffic scene with damaged pavement.(3)Campus traffic scene modeling of damaged pavement based on digital twinningTo achieve data interaction between the physical world and the virtual world,this paper maps the real-world ROS mobile robot to a virtual scenario of damaged pavement.The TCP communication protocol is used to achieve real-time position and direction synchronization of the virtual mobile robot with the physical ROS mobile robot and data communication of damaged pavement image information.Base on this,the traffic scene system of the damaged pavement is designed,which mainly involves the login interface,the auxiliary function of the system and the virtual-real interaction module of the disease data,and the digital twin scene modeling of the damaged pavement traffic is realized.