| The application of driving assistance technology reduces traffic accidents and improves automobile safety.With the gradual improvement of intelligent assistant technology,human driver will share control with advanced driver assistant system,which form a cooperative driving situation between human and automation.As one of the core technologies,driver-automation cooperative steering control has great potential in improving vehicle safety and subjective driving experience,while there are still many challenges in driver-automation cooperative steering control:(1)Driver-automation mostly forms a takeover-supervision relationship at present,which is easy to cause the driver’s over-reliance and situational awareness bias.How to ensure the driver’s participation in driving tasks under the premise of reducing the driving workload is a challenge;(2)The allocation of driving authority in driverautomation cooperative steering is directly related to vehicle safety,how to make a reasonable allocation criterion of driving authority is still need to research;(3)The essence of driver-automation cooperative steering is that two agents control vehicle steering together,how to describe the interaction relationship between driver and automation in cooperative steering control is a challenge;(4)The conflict between human driver and automation often arises in the driver-automation cooperative steering control,how to deal with the conflict is a core problem.In view of the problems above,this paper mainly focuses on driving authority allocation between drive and automation,driver-automation interaction relationship description and driver-automation cooperative steering control.In view of the problem that the driver’s low participation in driving task produces over-reliance and situational awareness bias,a driver-automation cooperative steering control structure,suit for L2 level intelligent vehicle,is designed to online optimize of the driving authority.This structure has the advantages of ensuring the real-time driving task participation of drivers.Under this structure,a novel cooperative steering control method considering driver’s driving sovereignty,driver’s characteristics and driver-automation conflict is proposed.This method can adjust the allocation of steering driving authority in real time according to the degree of driver-automation intention inconsistency.Not only the driver’s control authority is guaranteed,but also the driver’s workload is reduced.In order to meet the requirements of the description of driver-automation interaction of L3 level intelligent vehicles,a cooperative steering control structure of asymmetric information interaction is designed in this chapter,which realizes the description of asymmetric transmission of control information between human and automation and the description of automation intervention in human control.Aiming at the problem of driving authority allocation and cooperative steering control of intelligent vehicles,a driver-automation driving authority allocation logic based on risk assessment is proposed.Combined with driving authority allocation logic,a driver-automation cooperative steering control method based on safety envelope of the vehicle is proposed.The simulation results show that the proposed control method not only can reduce the driver’s workload,but can enhance the driver’s control ability of the vehicle.Considering that driver-automation interaction relationship can reshapes the driver behavior characteristics,a driver behavior model suitable for the development of driver-automation cooperative steering system is proposed,which makes up for the lack of human driver adaptability and learning ability expression in the design of cooperative steering system.On this basis,a driver-automation cooperative steering control structure considering symmetrical information interaction conditions is proposed,which improves the expression of handling information transmission from automation to human driver,and the adjustment rules for personalized driver’s driving authority is obtained through comparative experiment.In order to overcome the difficulties in evaluation of driver-automation cooperative steering system,a comprehensive evaluation index including subjective and objective is designed for the evaluation of driver-automation cooperative steering system,and the proposed method and the evaluation index provide references for the driver-automation cooperative steering system design.In order to improve the defect of the automation strong interference when there exist conflicts between driver and automation in the cooperative steering control,a driver-automation decision making model integrating the game theory is proposed,which reconstructs the driver-automation interaction relationship under the condition of symmetrical information interaction.And the condition of realizing game equilibrium and the optimal control strategy of driver and automation under the condition of game equilibrium are derived.The comprehensive evaluation of the designed cooperative steering system is completed by driver-in-the-loop experiments.The evaluation results show that the designed game theory based driver-automation cooperative steering strategy improves the subjective acceptance of the driver for the cooperative steering system.In this paper,the driver-automation cooperative steering control structure is described clearly,and the detailed derivation process of driver-automation relationship and control strategy is given.The comprehensive evaluation index of the driver-automation cooperative steering system including subjective and objective evaluation is designed,and the feasibility of the proposed driver-automation cooperative steering control schemes are tested and evaluated by using veDYNA dynamics simulation software and driver-in-the-loop experimental platform.The results show that the proposed cooperative steering control structure,driverautomation relationship description and comprehensive evaluation index have certain reference significance for the future research of driver-automation cooperative steering control.In the future,there will be further in-depth research in this paper: the description of driver-automation interaction based on diverse sensing and interaction tools,personalized driver’s driving intention understanding and driving behavior prediction. |
