Analysis And Identification Of Emotional Influence On The Dynamic Evolution Of Automobile Driving Propensity

Driving Emotion and driving propensity are both hot topics in the traffic field in recent years.With the rapid increase of the number of motor vehicles and the drivers,the Human-Car-Environment system is in serious conflict,traffic accidents happen frequently.Therefore,how to analyze the driver’s driving behavior and improve the safe driving assistance system has become the most priority in the field of transportation research.Drivers are the main participants of traffic with cognitive activities and vehicle handling ability,and play an important role in all aspects of the traffic system.Driving Emotion and driving propensity are the key factors to ensure drivers’ driving safety,the typical driving emotions include happiness,sadness,anger,anxiety and so on,while the driving propensity includes the conservative type,the ordinary type and the aggressive type.In this paper,the driving data under different emotions are analyzed,and the emotion inducing experiment,driving simulation experiment and human factors experiment are designed to obtain multi-source data,based on the theory of driving propensity and the deep learning model,the driving propensity is dynamically analyzed and identified,and the corresponding conclusions are drawn,it is of great significance to perfect the personalized early warning system of driving safety and the assistant safety driving system.Firstly,the information was collected by using the driving propensity questionnaire,the multi-person interactive driving simulator,the emotion maintenance time record scale,and the Psy LAB human factors engineering experiment system,so as to obtain the experimental data of drivers under different emotions,the data of driving behavior and human factors are analyzed and processed.Secondly,the quantitative expression model of driving propensity based on paired sample t-test and entropy method is established,with the help of variance analysis and box graph,the quantitative driving propensity is analyzed dynamically,and the dynamic transfer law of driving propensity under different typical driving emotions is defined,and the dynamic transfer probability matrix is established to verify it,the Identification Model of driving propensity is established for all experimental samples.After extracting the characteristic parameters by principal component analysis,the driving data with time series is input into the model.finally,a dynamic identification model of driving propensity which takes the intensity of driving emotion into account is proposed,taking the experimenter as the object of study,BPTT algorithm is used for short-term prediction of driving propensity in anger,and simple Bayesian cognitive science is used for real-time identification of driving propensity,the propensity of driving propensity changing with driving emotion intensity was analyzed.The results show that the model can effectively express the driver’s driving propensity,and can be used to analyze the change of driver’s driving propensity under different driving emotions.At the same time,the driving propensity identification model based on the whole sample and considering the degree of emotion intensity can accurately identify the driving propensity under different driving emotions and the dynamic change of driving propensity during continuous driving,the accuracy of identification is high.The research in this paper can enrich the theory of driver’s driving propensity,and is of great significance to the realization of driving safety warning and the improvement of vehicle safety and the assistant driving system.