Variable Structure Control Method For Human-vehicle Cooperative Driving

Semi-automatic driving is a stage that must be experienced in the development of intelligent driving technology.The main feature of semi-automatic driving is the sharing of driving rights in the case of that human-machine cooperative driving,which both the human driver(hereinafter referred to as human)and the vehicle automatic driving control machine(hereinafter referred to as machine)have the control right of the vehicle.How to effectively switch and coordinate control right between human and machine is the primary problem in achieving semi-automatic driving.In order to study the characteristics of human and machine interaction with traffic system under traffic scene,this thesis first constructs a human-machine cooperative driving system model.The model consists of a traffic scene model,a vehicle automatic driving controller model and a driver model.In this thesis,full velocity difference model is used as the traffic model,vehicle kinematics model is used as the vehicle automatic driving controller,and pseudo-mechanical model is used as the driver model.In order to realize the switching between the driver and the automatic driving controller,a sliding mode controller system based on stabilizing-regions is proposed,and a variable structure control method corresponding to human-machine cooperative driving is also designed.Firstly,according to the characteristics of the human-machine cooperative driving system model and the corresponding parameter regions,the stability regions of the driver control model and vehicle automatic driving controller model are obtained.On this basis,the driving evaluation method of the multi-parameter stability domain baesd on human-machine cooperative is completed.Secondly,according to the control model,the corresponding control quantities before and after the switching are obtained,and the switching function is designed to complete the switching transfer of the driving right between the human and the machine.Aiming at the problem that the sliding mode controller system based on stabilizingregions has insufficient smoothness in the driving right transition,this thesis proposes a sliding mode controller system based on smoothing.The design of the sliding surface is carried out according to the deviation of the driving system.The exponential approach law is used to optimize the switching process,and then the stability-based switching function is optimized by increasing the intermediate amount.By using the critical stability control quantity obtained by the sliding mode controller based on stabilizing-regions,the control quantities are optimized and improved by increasing the intermediate quantity,so as to optimize the switching process and complete the design of the overall sliding mode controller.The simulation experiments show that the two controller methods are feasible and effective under the four different initial typical conditions,which can be used to switch the driving right in the process of human-machine interaction.By camparing the results of the two methods,it is verified that the second method has better smoothness and stronger stability.