Research On The Method Of Intelligent Vehicle Following Preceding Vehicle In The Unstructured Environment

It is necessary to obtain accurate state information of the preceding vehicle to achieve autonomous following in unstructured environment.Accurate status information of the preceding vehicle is the basis for lateral and vertical control of the following vehicle.The overall framework of autonomous following is built based on the V2 V in the paper.And the method of acquiring the state information of preceding vehicle is put forward,considering the communication loss.Based on the historical position information of the preceding vehicle,the desired path of the following vehicle is obtained,as the constraint of the lateral movement of the following vehicle in lateral control.In vertical control,vehicle spacing and velocity control are based on the current vehicle spacing and speed velocity between the two vehicles.Based on the principle of leader-follower method and virtual structure method,a method is proposed to obtain the desired path using the historical position information of the preceding vehicle.Firstly,based on the unscented Kalman Filtering algorithm,the preceding vehicle state information is filtered and estimated to eliminate the influence of the communication loss.Then,based on different path model,the least square method is used to fit the location of the preceding vehicle to get its historical path.Finally,based on the bicyclemodel and the pure pursuit algorithm,a trajectory is obtained from rear vehicle and converges to the preceding vehicle's historical path.And make the trajectory as the desired path of the rear vehicle,so the desired front wheel angle of the rear vehicle can be obtained,and tracking the preceding vehicle's history path is implemented.In the longitudinal control of follower vehicle,a longitudinal control method based on sliding mode control theory and vehicle spacing error model are proposed.Firstly,the characteristics of different vehicle spacing strategies are analyzed,and the Constant Spacing Strategy is selected.Then,the sliding mode controller of vehicle spacing is designed based on exponential approach rate and shift reaching law,and the chattering phenomenon near the equilibrium position is eliminated by the combination of two reaching laws.After that,considering the maximum velocity of the vehicle and the maximum acceleration and deceleration,the vehicle spacing controller is improved.Finally,considering the limitation of the vehicle preview characteristics,the spacing controller is further improved,when the spacing error is large,the desired acceleration of the rear vehicle is obtained by velocity planning,to make spacing error reduced to a certain range,and then start the sliding mode controller.Finally,based on the V-REP simulation platform and the real vehicle platform,the proposed control method of autonomous following preceding vehicle is verified experimentally.The result shows that using the method proposed in this paper,the follower vehicle can follow the preceding vehicle automatically in straight road,left / right turning Road,U curve road and S curve road in the unstructured environment.On a rugged road,the maximum speed reached 12m/s,when the distance is controlled in the range of less than 30 m.