Driver-automation Cooperative Control Approach In Avoiding Obstacles For Intelligent Vehicle

Automatic driving plays an important role in reducing the traffic accidents caused by the driver's error,which attracts more and more attention in recent years.With the development of science and technology,driverless driving technology has made great progress,but there are still many technical limitations that make it difficult to cope with the complex and changeable traffic environment,particularly in multiple obstacles condition.Therefore,driver-automation cooperative steering has become the focus of attention in recent years,which not only refers to the automatic controller design but also involves authority allocation strategy.In this paper,in view of obstacle avoidance conditions,a reasonable and effective automatic controller and authority allocation strategy is constructed to achieve driver-automation cooperative steering,as shown below:Firstly,to design the automatic controller,based on newton's second law,the vehicle point-mass model and three-degree-of-freedom(3-DOF)dynamic model are developed under the necessary assumptions,which can describe the kinematics and dynamic characteristics of the vehicle,respectively.The road and obstacle potential functions are constructed based on the principle of artificial potential field method.The construction of vehicle model and potential function lays the foundation for the subsequent design of the automatic controller.Secondly,the automatic controller,including path planning and tracking modules,is designed based on model predictive control(MPC).A path planning module for static obstacles is designed by deriving the stability constraints and taking the point mass model as the prediction model.In addition,the path tracking module is constructed based on the linearized 3-DOF dynamic model,and the constraints of vehicle actuator are considered.Automatic control of vehicle is realized by both path planning and tracking modules.Thirdly,a concept of space collision risk(SCR)is introduced for assessing the vehicle safety state,and the effects of relative distance and angle on SCR is analyzed in detail.The weighted summation authority allocation strategy is developed based on SCR,to blend steering instructions generated from driver and automatic controller.In addition,to mimic the driver's character in operating the vehicle,an optimal preview lateral acceleration driver model is established based on preview-follow theory.Finally,the potential field function for dynamic obstacles is constructed,the position and velocity of vehicle relative to obstacles are considered comprehensively,and the potential field function for two-lane road with desired lane and overtaking lane is established.Based on the potential field function,the automatic controller for obstacle avoidance,by means of lane changing and overtaking,is constructed.And the effectiveness of proposed authority allocation strategy is verified in the situation of dynamic vehicle obstacles.In this paper,the driver-automation cooperative steering control approach under obstacle avoidance condition is studied,and the driver-automation cooperative steering system is built based on the CarSim-Simulink co-simulation platform to verify the effectiveness of the above control method and authority allocation strategy in driver-automation cooperative steering process.