Human-Vehicle Shared Steering Control Considering Driving Ability Compensation

In recent years,autonomous driving technology has gradually matured.However,the current autonomous driving technology cannot reach the level of fully autonomy in a short period of time.More or less,it requires the driver and the vehicle system work together to control the car together.It requires the division of labor and the vehicle system to control the car.As matter of fact,the matter of shared autonomy has turned into a aroused general interest of intelligent automobile research area,and has aroused numerous investigators’ care.one out of a multitude problems of shared autonomy is to coordinate and distribute the control rights of man and machine when it’s on manvehicle simultaneously control.It’s essential to warrant that the driving assistance system is safe and effective for the driver’s assistance control,and to bring into ample play to the system’s effect,but also to lower the confrontation between the system and the driver,ensure that the driver has sufficient driving participation experience.Aiming at the collaborative problem of human-vehicle participation in driving tasks,investigators have proposed a parallel control mode of man-machine in the loop at the same time,namely,human-vehicle sharing control(HVSC).Shared control schemes can be divided into two categories according to different modes of action:haptic feedback shared control and input mixed shared control.Contrasting with overseas,researches on input mixed shared control optimized for different types of drivers are less in China.This paper adopts the input mixed shared control scheme,and proposes a shared steering control method for human-vehicle sharing based on the compensation of driving ability of different types of drivers.First of all,a preview driver model is established.According to the driver characteristics of preview-follow theory,the driver model is divided into two types: skilled driver and novice driver,and according to the influence of road curvature on the preview time of the driver,an adaptive adjustment method for the preview time of different types of driver models is proposed.Secondly,An automobile sharing steering control system based on Model Predictive Control(MPC)is designed.In the cost function,the lateral error and heading angle deviation are set as the state cost,and the steering wheel angle increment is set as the control cost.A fuzzy control-based weight distribution method for human-vehicle driving is proposed and adopted,considering the change of road curvature and vehicle speed,select the appropriate lateral error standard factor,and then adaptively adjust the weight distribution coefficient of human and vehicle driving to make a reasonable dynamic allocation of the driving weight of human-vehicles,which achieves a good vehicle driving effect while minimizing the system’s intervention to the driver.Finally,the designed automobile steering assist system was simulated under various working conditions based on MATLAB/Simulink and Carsim simulation environment.From the experimental results,the designed car steering assist system can play a good assisting effect for two types of drivers;the steering assist system can make the novice driver close to the driving effect of the skilled driver to a certain extent;when the driver can better control the vehicle,the steering assist system has little effect on the driver,so that the driver maintains a high degree of control over the vehicle;as the driving difficulty increases,the driver’s driving effect becomes worse,and the steering assist system intervention can be appropriately increased according to the driver’s control effect,so that the vehicle ultimately maintains better tracking accuracy and steering stability.