Research On Rolling Optimization Control For Driver-Automation Cooperation Steering Of Intelligent Vehicle

With the development of automobile control,the shared steering control concerned with the cooperative work of driver and automatic system has become a new research hotspot.When a human driver shares steering control with an automatic system,the relationship between the driver and the electronic control system is no longer decision and execution.The participation of the control system directly affects the driver's driving authority and the security of the system.Solving the problem of lane keeping or obstacle avoidance,the shared steering control is a concern in this paper when it fulfills the driver's sufficient driving authority and safely completes the steering task within the road.When considering drivers and automated systems as two agents,how to balance the driving weights of both is also a concern in this article.At the switching control between the driver and the automatic system,how to achieve flexible switching is also the focus of this paper This paper research and analysis as follows.The scheme that mixing the independent outputs of The driver and the automatic system to get the final order is more complex,and will cause the driver of the vehicle to the feedback of poor feelings.So the problem of sharing steering is transformed into an optimization problem so as to obtain the front wheel angle directly in this paper.A constrained model predictive control method is used to design a shared steering control controller.Road safety constraints considering vehicle shape and stability safety constraints are transformed into constraints in model predictive control.The driver's steering intention is considered as the main optimization target in the model predictive control.The simulations of different conditions verify the result as far as possible to ensure the driver's steering intention in a safe case.In order to balance the driving weights of the driver and automatic system,we analyze the shared steering problem and design the controller from the dangerous level.On the basis of the constraint model predictive control scheme,the control objective of the automatic controller following the road central line is added,and the weight coefficient of the automatic driving is determined by fuzzy rules according to the road danger level and the driver's danger level.The current and predicted location information of vehicle and road is used to determine the road danger level,and the driver's danger level is determined by the predicted steering angle and the current driver's steering command.In addition to changing the driving weight through danger assessment,this paper also analyzes the driving authority distribution problem of the driver and the automatic controller in the prediction horizon.Based on the constrained model predictive control method,variable weights in predictive horizon scheme is proposed.We select the linear piecewise functions as the driver and automatic system weights functions.Driving authority is shifted from driver's command to autopilot target in the prediction horizon.In order to compare and evaluate the performance of different schemes,the control intervention rate and risk rate are given.Finally,through the simulation of the safety and dangerous conditions compared,the intervention rate and risk rate are statistically analyzed.The results show that the two improved schemes reduce the risk rate.When discussing the possible instability of switching in switch control,the driving weight function of the driver and the automatic system is obtained in real time based on the prediction window,and the takeover task is completed through solving the optimization problem.In this paper,two kinds of takeover procedure of the driver to the automatic system and the automatic system to the driver are given.