Study On Road Area Risk Assessment And Prediction Method Based On Inter-Vehicle Relationship

With the development of the social economy,the car has entered thousands of households,bringing great convenience to our travel.It has brought convenience to our travel,but it has also brought many problems.The most notable one is the traffic safety problem.Traffic accidents bring huge losses to the safety and property safety of the people every year.How to alleviate traffic safety has become a difficult problem for governments and researchers.With the advent of the Intelligent Transportation System(ITS),there has been a major turnaround in the mitigation of road traffic problems.In particular,the development of driver assistance systems has greatly improved the safety level of driving.The real-time assessment and dynamic prediction of road area driving environment has always been one of the hotspots in the field of driving assistance system research,but the effects of evaluation and prediction have been affected by the validity of the calculation model.So we propose a method for real-time assessment and dynamic prediction of road area hazards based on inter-vehicle relationship theory.The research content of this paper is as follows,Firstly,we establish a model of road area inter-vehicle relationship.Through this model,the road area inter-vehicle relationship value can be obtained.The road area inter-vehicle relationship value is the main parameter of the road area hazard real-time assessment and dynamic prediction function,which is mainly determined by the vehicle health status and the related attributes between vehicles.Based on the Ising model,this paper establishes an inter-vehicle model considering the road traffic factors such as vehicle density,inter-vehicle distance and inter-vehicle relative speed in the target road area.This model is the basis for real-time hazard assessment and dynamic prediction of road areas.Secondly,we propose a road area hazard real-time assessment model based on inter-vehicle relationship.The model is divided into two parts: road area hazard real-time assessment and road area hazard real-time distribution calculation.We are inspired by the method of studying the magnet system in the Ising model.The road area is regarded as a two-dimensional plane,further divided into small grids,and then the assessment model of road area hazard is established by referring to the group energy formula in the Ising model;In addition,based on the value of the inter-vehicle relationship,the hazard values of each small grid are obtained in turn,and the dangerous distribution of the road area is reflected by the hazard value of each small grid.Thirdly,we propose a road area hazard dynamic prediction method based on real-time evaluation values,which is mainly divided into two parts: dynamic prediction of road area hazard and dynamic prediction of road area hazard distribution.The method dynamically predicts road area hazard values for several consecutive moments in the future based on the real-time evaluation values of the road regions in the most recent continuous time.At last,based on the Anylogic traffic simulation platform,the NS following model is used to simulate the road driving environment,and relevant traffic data is obtained to conduct experiments to verify the evaluation effect and prediction effect of the proposed method.The experimental results show that the proposed method can effectively track and reflect the changes of the driving environment in the road area,which is of great significance for the improvement of the performance of the driving assistance system.In addition,the results of the evaluation and prediction can also be used for traffic safety warning and path planning.