Research On Driver’s Visual Load Of Mountainous Expressway Based On Scene Change Information

In recent years,with the vigorous development of highway traffic,a large number of mountain expressways have appeared in the central and western regions of China.Due to the complex terrain and geological conditions of mountainous expressways,a large number of special sections composed of typical structures such as bridge and tunnel groups,tunnel groups,and tunnel-interchange small-spacing sections have been formed.The traffic environment information of this section is complex and changeable,and the driver ’s visual state is constantly changing,which will have a great impact on the driver’s life psychology,and it is easy to have traffic safety accidents caused by the coupling of people,vehicles,roads and environment.At present,domestic and foreign researchers have done a lot of research on the visual characteristics and visual information of drivers on mountainous expressways,mainly focusing on the visual characteristics of single structures(single tunnels,single bridges,single interchanges)and tunnel groups and the influence of traffic sign information on the visual characteristics of drivers.However,the comparative study on the visual load and visual information of drivers in typical scenes such as bridge and tunnel groups,tunnel groups,tunnel-interchange small spacing sections and ordinary subgrade sections in mountainous expressways and the research on safety assurance technology are still shallow.In this thesis,the visual load and visual information of drivers in four typical scenes of special sections of mountainous expressways are studied in depth,and corresponding safety measures are proposed.The main research contents and conclusions of this thesis are as follows.Through the real vehicle test,four typical scenarios of the driver in the special section of the mountainous expressway were collected,namely,the eye movement characteristic data,the driving speed data,the driving image data and the light intensity data of the bridge tunnel group,the tunnel group,the tunnel-interchange small spacing section and the ordinary roadbed section.Based on the real vehicle test data,four eye movement indicators of pupil diameter,fixation time ratio,average saccade time and blink frequency are selected.By comparing and analyzing the visual characteristics of drivers in different scenes of special sections of mountainous expressways,the bridge and tunnel group sections of bridges or superlarge bridges,tunnel group sections below 400 m of connecting sections and tunnels with spacing less than 400 m are defined as high-risk scene sections of mountainous expressways.The principal component analysis method and the entropy weight TOPSIS method were used to establish the calculation model and evaluation model of the driver’s visual load intensity in the special section of the mountainous expressway.The K-means clustering analysis was used to divide the classification standard of the driver ’s visual load intensity in the special section of the mountainous expressway,and the threshold range was determined.Using the data and images collected by the real vehicle test,the information entropy theory and image processing technology are used to establish the visual information quantification model of the special section of the mountainous expressway.The visual information of the bridge tunnel group,the tunnel group and the tunnel-interchange small spacing section is compared and analyzed.The results show that the relationship between the visual information amount of the driver in the special section of the mountainous expressway is: tunnel group > bridge tunnel group > tunnel-interchange small spacing section > ordinary roadbed,and the information amount of the high-risk scene section is greater than other sections of the same scene.On this basis,the author analyzes the relationship between the driver ’s visual load intensity,visual information and driving speed in the bridge and tunnel group,tunnel group and tunnel-interchange small spacing section of mountainous expressway,and reveals the inherent law between the three.The results show that the driver ’s visual load in the special section of mountainous expressway is positively correlated with the visual information and negatively correlated with the driving speed.Based on the support vector regression(SVR)model,the regression calculation model of driver ’s visual load intensity,visual information and driving speed in special sections of mountainous expressway is constructed.The validity of the driver ’s visual load intensity model and visual information quantification model in special sections of mountainous expressway is verified.Finally,the internal interaction mechanism among the three is discussed.Finally,combined with the characteristics of the traffic environment and the driver’s visual characteristics of the special section of the mountainous expressway,the corresponding driving safety guarantee technology is proposed for the high-risk scene section and the ordinary roadbed section from the aspects of light environment and road traffic environment information,and the safety guarantee measures are evaluated by using the scale model and UC-win / Road simulation software.The simulation results show that the safety guarantee measures proposed in this thesis for high-risk scene sections and ordinary subgrade sections have a good effect on improving the driving safety level of special sections of mountain expressways to a certain extent,and also provide a theoretical basis and reference basis for the research of safety guarantee technology of this section.