Research And Application Of Man-machine Weight Allocation Strategy In Lateral Motion Control

With the increase of vehicle amount,the problem of traffic accident is becoming more and more serious.The application of various advanced driving assistant systems has become an effective way to solve this problem,and also the way leading to driverless cars.However,with the further application of advanced driving assistant system,how to deal with the interaction between driver and assistant system has become an open issue to be solved.Therefore,this paper focuses on the controller design in lateral motion control and the weight allocation strategy design between driver and assistant system.Then,the assistant effect of weight allocation strategy on drivers is analyzed.Lane keeping is the main measure of lateral motion control.Firstly,this paper gives a brief description of lane keeping assistance system and weight allocation strategy,reviews the research progress and academic achievements of domestic and foreign researchers in these two fields.After that,the vehicle-road model and steering system model are built.By analyzing the actual steering behavior of driver,the neural network driver model based on the two-point driver preview theory is established.The rationality of the driver model was verified through the simulation experiment and the comparing with the optimal preview curvature model.The lateral controller is designed on the control theory of model predictive control,and considers the constraints in the process of driving(such as stability constraints,the actuator control constraints)and robustness requirements.For the unknown parameters of electric power steering system,the sliding mode observer is designed.Finally,the proposed controller is verified by simulates in different tire-road friction coefficient and different speed.A weight allocation strategy is proposed to solve the problem of control authority conflict between driver and assistant system in lateral motion control.Firstly,the cooperative states between driver and assistant system are divided,and the influence of weight coefficient on lateral motion control effect and driver's subjective feeling is explored.Then,the risk level of lane deviation is evaluated based on preview distance to lane center,and the weight allocation strategy is designed when driver and assistant system are in the control loop at the same time.After that,authority transfer strategy is designed according to the torque difference between driver and assistant system.The strategy is a revision of the weight allocation results.Finally,the proposed weight allocation strategy is simulated and verified.It's effectiveness in improving the accuracy of lateral control,reducing the workload of driver and correcting the misoperation behavior of driver is also analyzed.At last,the hardware-in-loop experiment platform of lateral motion control is established based on CarSim/Labview.The effectiveness of the weight allocation strategy and the lateral controller proposed in this paper is verified as well.