Research On The Running Characteristics Of Undercurrent In The Light Environment At Night

With the rapid increase of car ownership,the problem of road traffic safety is becoming more and more serious.Nighttime is a period of high incidence of traffic accidents and the degree of injury is more serious.The root of the impact on traffic safety lies in the change of driver's visual characteristics and traffic flow operation characteristics.Therefore,in the light environment at night,it is of great significance to combine the research on single vehicle operation and vehicle flow operation characteristics.First of all,on the basis of relevant research at home and abroad,this paper combs the driver's visual characteristics and vehicle flow operation parameters,designs the driver's recognition distance real vehicle experiment and the use of UAV aerial video experiment,and collects the road information,processes the vehicle flow video data through the Tracker software,and obtains the actual vehicle flow operation parameter data.Then,in the aspect of single vehicle operation,carry out the research on the driver's recognition distance of different illuminance and different road sections,draw a conclusion through the comparison between day and night,at the same time,use SPSS to establish the nonlinear regression model of recognition distance with illuminance,and study the running speed characteristics of vehicles on the same road section with different illuminance and different roads with the same illuminance.On this basis,in the aspect of running characteristics of traffic flow,use Matlab to draw scatter diagram to compare the speed and acceleration of traffic flow in daytime and night,and study the different distribution of speed and acceleration of different traffic flow when the illumination changes,as well as the change state of speed,headway and headway of the same traffic flow.Finally,continue the previous research,through the principal component analysis of multiple factors of night driving environment,with illuminance,traffic flow,number of lanes and whether there is a central isolation belt as the input variable,average speed as the output variable,establish BP neural network through Matlab,and verify the reliability and effectiveness of the network model,and train the network model with error satisfying conditions,for the night The theoretical basis is provided for the operation of inter vehicle flow.