Analysis Of The Interaction Effect Of Driver Behavior And Emotion In Car-following Driving

With the development of science and technology and the introduction of related policies,the intelligent networked vehicle industry has become a long-term development element of China’s science and technology.And one of the key elements of the intelligent networked vehicle is to make the vehicle simulate the emotion and cognition of human driving in the operation process,understand itself and the surrounding driving situation,and be able to make planning decisions intelligently and independently.Therefore,in order to explore the interaction between emotion and behavior in following driving,and to establish a runnable emotion-behavior model for intelligent vehicle systems,this paper analyzes the effects of different emotion-behavior interactions of drivers in the driving process,as follows.Firstly,starting from the basic research of emotion,the emotion induction experiment and the simulated driving experiment were developed.The data of emotional characteristics,physiological characteristics,simulated driving data and driving style data of the experimenter under different emotions were collected through the experiment.Next,the physiological data were preprocessed and assigned weights using a combination of principal component analysis and entropy method of assigning weights to the physiological characteristics data.Subsequently,ANOVA was performed on the driving behavior data under different emotions to explore the behavioral characteristics indicators with significant differences under different emotions,and two-by-two analysis was conducted to find the differences between the driving behavior data under different emotions according to the horizontal and vertical motion directions.Factor analysis was performed on the driving behavior data,and the extracted principal components were named as lateral and longitudinal motion factors according to motion characteristics,and clustering was performed using the K-means method to classify the driving behavior data under different emotions into low-risk,medium-risk and high-risk driving behaviors,and the clustered driving risk behavior data were verified using a support vector machine(SVM).Finally,the physiological characteristics,driving behavior,emotional characteristics and driving style data were analyzed,and a structural equation model(SEM)was used to construct a model of the behavioral effects of the emotional state of driving,and the effects of the six basic emotions of happiness,anger,sadness,fear,disgust and surprise on driving behavior and between physiological indicators and driving style were quantitatively analyzed.A structural equation model was also used to construct a model of the effect of emotion on driving behavior,and the effect of driving behavior on the six basic emotions,as well as the effect of physiological indicators and driving style,was quantitatively analyzed.The results of the study showed that the effect of emotion on behavior was greater than the effect of behavior on emotion.In particular,emotions tend to control physiological responses through cognitive activities,and this process is more influential.Behavior,on the other hand,tends to regulate emotion through event evaluation,which is a weak influence.The research content of this paper provides the theoretical basis for the proposed human-vehicle and vehicle-vehicle interaction technology in the unmanned driving technology and vehicle networking environment,which is of significance for realizing the unmanned driving of vehicles and improving road safety.