Research On Path Tracking Control Of Intelligent Vehicle Based On Human-Machine Co-Driving

Human-machine co-driving denotes a multidimensional collaboration between the human driver and the automated system,which shares control of the vehicle and accomplishes the driving task collaboratively.This approach can solve the safety and cost dilemma of high-level intelligent vehicles,which is a new approach and idea to improve the intelligence level of vehicles.A shared control approach that ensures the driver is in the loop is a fundamental prerequisite for human-machine co-driving.This paper focuses on human-machine co-driving vehicles,aiming to achieve high precision safe path tracking control.At present,similar studies suffer from constrained path-tracking performance and insufficiently rational methods of allocating driving rights.In order to overcome the above problems,this paper conducts a study from the tandem and parallel steering-sharing human-machine co-driving framework.The core issues of human-machine co-driving system are analyzed,and the functional objectives of human-machine co-driving are further clarified.Based on the above,aiming at the tandem and parallel steering-sharing human-machine co-driving framework,the overall architecture of the existing intelligent vehicle is expanded,and the overall architecture of the human-machine co-driving system is designed.Subsystems such as driver monitoring and driving rights allocation are added,and their working principles and implementations are analyzed.Aiming at the driver-in-the-loop control problem,a tandem steering-sharing humanmachine co-driving control method is proposed based on model predictive control.The system state and output of a human-machine co-driving vehicle were predicted by analyzing the vehicle dynamics and kinematics.Considering the road boundary constraint and actuator saturation constraint of the actual vehicle,etc.,the optimization function of the human-machine co-driving vehicle is established.The two-point preview driver model is used to simulate human drivers,which is simple and efficient to realize human-machine co-driving.In order to better deal with vehicle uncertainty and driver state fluctuations,a parallel human-machine co-driving control method based on constraints-following control is proposed.The system uncertainty is decomposed for the underdrive characteristics of front steering vehicles,and the path-tracking equality constraint and road boundary inequality constraint are integrated.A path tracking robust controller is designed based on constraints-following control,and the system stability of this controller working alone is demonstrated.Taking into account the path tracking accuracy,the degree of human-machine conflict and the driver’s driving capability,a numerical model for the allocation of driving rights is constructed.Humanmachine co-driving is achieved by adjusting the control authority of the controller and the driver.A joint Carsim-Matlab/Simulink simulation platform is established,and the simulation verification of the tandem and parallel steering-sharing human-machine co-driving control method is conducted.The path following accuracy and driving safety of the controller were verified,and then simulations were conducted for the double lane change path and the snake path respectively for different driving states of the driver.It is verified that the proposed method can achieve high-precision safe path tracking for human-machine co-driving vehicles,and weaken the negative influence of drivers.