Study On Flexible Control Transition For Man-Machine Shared Driving Under Human-Vehicle-Road Interaction

Drivers' misconduct is the main cause of traffic accidents.With the development of intelligent driving technology,the on-board driving system has become the second driving agent which could help human drivers in vehicle control,correcting occasional misbehaviors of human drivers so as to improve driving safety.Under the man-machine shared driving mode,the driving control authority transition mechanism plays a decisive role in coordinating the man-machine relationship.To fully utilize the current intelligent driving system for the improvement of driving safety,this paper adopted the vehicle traffic as the research object,and proposed a unified quantitative model of driving risks to explore intrinsic response characteristics of driving risks in the driver's driving process.Additionally,the average response characteristics of drivers were adopted as the behavioral benchmark for drivers and to put forward the flexible control transition model(FCTM)based on the response deviation.In assessing the on-road traffic risk,the driving risk field was used as the theoretical framework.In light of the static electric field in physics,a scalable and transferable electric-like risk field model was proposed to measure driving risk.In this model,the inherent risk-related characteristics of a traffic unit is described by the risk volume,and the impact of motion state and the features in traffic scene on the risk situation is characterized by the intensity of risk distribution.In analyzing characteristics of drivers' response to risks,road experiments in carfollowing and side-car cut-in scenarios were conducted to extract the characteristics in longitudinal risk-response from the driving data of 24 drivers.Based on the extracted characteristics,a human-like car-following model,namely the risk-response driver model(RRDM),was proposed.The simulation results suggest that the car-following model based on the characteristics in longitudinal risk-response can effectively reproduce drivers' driving behavior.Next,according to preliminary exploration results in longitudinal scenarios,more driving data were extracted from HighD natural driving dataset so as to comprehensively analyze the omni-directional risk-response in drivers' car following and lane changing process.The analysis results show that drivers follow specific and quantifiable risk-response modes in coping with different driving tasks,and these modes are recoverable among driver groups.Regarding the FCTM design,the average risk-response characteristics of drivers in the process of safe driving are taken as the benchmark for driving behaviors.According to the degree of response deviation,the driving control is switched to the driving system to modify drivers' irrational behaviors.Finally,the simulation scenarios of the front-car emergency braking and side-car cut-in were designed.The hardwarein-loop simulation experiment platform was built,and then the man-machine shared driving and control system based on the platform was established.Through the simulation experiment,the performance of flexible control transition model for manmachine shared driving in improving the driving safety was verified.