| In order to explore the correlation between driver visual behavior and hazard perception abilities in navigation environments,traffic scenes in navigation environments were divided into ordinary through traffic,highway,lane change,turn,and special road traffic scenes.Combining driver hazard perception experiments,the correlation between driver visual behavior and hazard perception abilities in different traffic scenarios with or without navigation environments was analyzed;The influence of navigation environment on driver’s visual behavior and the correlation between visual behavior and hazard perception ability are analyzed;Based on the analysis results,a prediction model for driver’s hazard perception ability in a navigation environment is constructed;Based on the research results,suggestions are proposed to enhance the driver’s ability to identify potential hazards on the road,improving the driver’s hazard detection system,and effectively utilizing the driver’s hazard perception skills.The main research content is as follows:Firstly,through correlation analysis and variance analysis,the correlation between driver visual behavior and hazard perception was explored from four types of eye movement: saccade,fixation,eye activity,and area of interest attention rate.The results showed that the number of times the driver’s hazard perception ability index did not recognize hazard has a significant negative correlation with the average fixation time of the fixation visual behavior,and has a significant correlation with the average fixation time of the fixation visual behavior There is a significant negative correlation between the eye activity index level and the attention rate index in the front area of the attention rate index;The study found a strong correlation between the average fixation count and the number of driver hazard recognition errors;and finds a notable correlation between the driver’s correct recognition of hazards and their average duration of visual fixation.Additionally,the study reveals a significant relationship between the average duration of saccades and the driver’s level of attention to the straight ahead area.In the same traffic scenario,the lower the driver’s fixation time and scanning time,the lower the attention rate in the area directly ahead,and the more times the driver correctly recognizes hazards.There was a notable correlation between the average scanning angle and the number of hazard recognition errors;The average number of saccades and vertical eye activity indicators were significantly correlated with the number of hazard recognition errors.Secondly,under different traffic scenarios,the differences between driver visual behavior and hazard perception in the presence or absence of navigation environments were compared using paired sample t-tests.The results showed that in highway traffic scenarios,there were significant differences in drivers’ horizontal eye activity indicators and attention rates in the immediate ahead area,and there were significant differences in the number of times hazards were not recognized;In steering traffic scenes,there is a significant difference in the average scanning time of drivers’ scanning visual behavior;In lane changing traffic scenarios,there is a significant difference in the number of times drivers fail to identify hazards and the number of times they correctly identify hazards;In special road traffic scenarios,the study revealed a notable disparity in the frequency of driver’s correctly identify hazards.Finally,based on the correlation between driver’s visual behavior and hazard perception ability and the impact of navigation environment on driver’s visual behavior,a prediction model for driver’s hazard perception in navigation environment was established by combining principal component analysis and generalized linear model,and the degree of fitting of the model was tested and the effectiveness of the model was verified.The research content and results of this thesis have important theoretical significance for improving the visual behavior system of driver’s hazard perception;It has guidance and reference value for the development direction of driving assistance information provided by vehicle auxiliary equipment and artificial intelligence in the future. |