Study Of Human-Machine Interaction And Visual Distraction Characteristics In Vehicle Full Touch Mode

Touch HMI(Human Machine Interaction)is a new technology in the field of vehicle human-machine interaction.In terms of driver operation,the operation of touch HMI is actually similar to the operation of a tablet computer,which changes the traditional HMI operation behaviour pattern of the driver.The visual characteristics of the driver when performing HMI tasks with a full-touch screen,as well as individual driver differences,have not yet been studied in depth and are closely related to the risks of vehicle operation caused by full-touch HMI.To address the above issues,we designed and built a touch HMI vehicle experiment platform,selected a typical HMI operation task,and conducted a real-world driving experiment on 29 subjects in a real road environment,collecting data related to the operation of the humanvehicle system at different speeds when the driver used the touch HMI to complete a specified sub-task.The 3480 sets of touch HMI subtask operation data were analysed to investigate the characteristics of the impact of vehicle speed and task complexity on the visual distraction characteristics of the subtask process,and the distraction style idea was proposed to discuss the individual differences of drivers,and a driver visual distraction duration prediction model was established,with the main research content as follows:(1)The effects of vehicle speed and task complexity on drivers’ visual distraction characteristics were explored based on three subtask data: air conditioning operation,phone calls and navigation.As the complexity of the subtasks increased,drivers needed more operational steps and attention resources to complete the tasks,which led to an increase in the total visual distraction time,number of visual distractions and maximum visual distraction time;as the speed increased,the total visual distraction time,average visual distraction time and maximum visual distraction time decreased due to risk compensation.In terms of numerical magnitude,some of the visual distraction duration parameters generated by completing subtasks exceeded the 2-second threshold,15-second principle,etc.identified in existing studies,posing a greater safety risk.(2)With reference to existing research systems on driving styles,this study proposes a method for describing individual driver differences using visual distraction parameters and uses the concept of distraction style to describe such individual differences.The cohesive hierarchical clustering method was used to classify drivers into three categories: cautious,normal and aggressive,and to further analyse the differences in visual distraction characteristics of drivers with different styles.The results showed that cautious drivers tended to have multiple short-duration visual distraction patterns to complete subtasks,and aggressive drivers tended to have fewer long-duration distraction patterns to complete subtasks.(3)With reference to the computer interface human-computer interaction KLM-GOMS model,a method for predicting the total duration of subtask visual distractions was proposed to achieve a prediction of the duration of visual distractions during subtask operation of the touch HMI,with a prediction accuracy of 84.3%.Based on this,the prediction model was refined by introducing distraction styles,and the results showed that the maximum improvement in the prediction accuracy of the model was 5% compared to that without considering distraction styles.