A Study On Rainy Highway Speed Limit Considering Driver’s Visual Perception Characteristics

With the rapid development of Chinese social economy,Chinese highway network construction is also developing rapidly,the total mileage of highways and expressways open to traffic ranks first in the world.Highway transportation network is convenient for people’s life,but also due to the influence of bad weather such as rainy days lead to frequent traffic accidents,resulting in major economic losses and personal casualties.In this paper,on the basis of the domestic and foreign literature on the speed limit of rainy road and drivers’ visual recognition ability,based on the visual recognition of drivers’ signs on rainy road speed control scheme is studied.First of all,this paper analyzes rainy weather conditions and speed: visibility to drivers,based on this put forward based on the Fisher linear discriminant,bad as in rainy day,watching the speed and validity as input variables,to vertical vision discrimination as a result,the identification or miss as the classification standard of considering driver visibility sign model of the speed limit.Furthermore,the rain identification model based on Alexnet neural network and the rain severity quantification model based on Brenner gradient evaluation function were established by analyzing the rainfall grade and the minimum visibility range.Analyze the driving field of vision and establish the vertical field model.This paper analyzes the range of vertical visual field required by drivers to accurately identify identification plates,determines the threshold value of vertical visual field and puts forward the basis for dividing vertical visual field.Secondly,the real vehicle road calibration test and simulated rainy day scene calibration test were designed in rainy day environment.The identification distance required for each rainfall level under 50km/h speed of real vehicle road test was calibrated to simulate rainy day scene,and the setting of weather environment parameters was determined.At the same time,in order to avoid the traffic scene screen displayed by the simulator being reduced,The vertical field adjustment coefficient was determined according to the pixel ratio of the sign in the eye tracker recorded video at the same position in the real vehicle test and the simulated calibration test.Based on this,the simulated rainy day highway test scene was built,the test scheme was designed,and the data preprocessing method was proposed.Finally,the relationship between rainfall grade,minimum visible range and the severity of rainy day was analyzed.The relationship between drivers’ fixation interest area,the proportion of fixation points in each fixation interest area,fixation validity,vertical field of vision,recognition distance and vertical field threshold,vehicle speed and the severity of rainy day was studied.Fisher linear discriminant was evaluated to compare the proposed rainy day road speed based on the visual recognition of driver’s sign and the traditional proposed rainy day road speed based on the parking sight distance.The results show that the traditional recommended speed based on parking sight distance is relatively high under the same severity of rainy days,and the speed limit control should be optimized in order to fully consider the perception of drivers on the premise of ensuring the safety of driving on rainy days.The research results provide a theoretical method for the correlation between drivers’ visual recognition characteristics and road traffic safety,and provide a theoretical basis for formulating the recommended speed value and the size design of traffic signs on rainy days,which has a certain practical significance.