| With the growth of population and the expansion of the international market,improving the efficiency of rail transit transport can transport more passengers and cargoes,meanwhile,responding to China’s policy of building strength in transport in the new era.Virtual coupling can improve the transportation efficiency of existing lines,which is an effective way to solve this problem.There are some key problems in the realization of virtual coupling.Firstly,the speed tracking error of the virtual coupling train should be reduced to realize the cooperative control of virtual coupling train formation.Secondly,the tracking distance of the virtual coupling train formation should be kept to solve the problem of collision avoidance.Finally,the stop accuracy of train formation should be solved.This paper proposes a scheme of virtual coupling train operation control by analyzing and understanding the concept of virtual coupling and combining the characteristics of the existing train operation control system.The main factors that affect virtual coupling train operation control are analyzed.Virtual coupling train formation control method based on generalized model predictive control is proposed to realize the cooperative control of virtual coupling train formation.An anti-collision control method based on variable parameter artificial potential field for virtual coupling train formation is proposed to realize the anti-collision of virtual coupling train formation.The generalized model predictive control and variable parameters are fused,and a fusion control algorithm of the virtual coupling train is proposed,which meets the precision requirements of the control algorithm in the virtual coupling train scheme.The main research works are as follows:(1)By analyzing the operation scenarios of the virtual coupling train,and drawing on the leader-follower formation method which is widely used and easy to implement in multi-agent formation,this thesis proposes a virtual coupling scheme based on the local dual leader-follower model which considers both the control precision and the implementation difficulty.Besides,the control precision that needs to be satisfied to realize the scheme is analyzed.(2)In order to realize the cooperative control of the virtual coupling train,the generalized model predictive control algorithm based on least square parameter identification is proposed.The actual operation data of high-speed train collected on site is used to verify the cooperative control effect of virtual coupling trains.The results show that the maximum speed tracking error of the cooperative control algorithm based on generalized model predictive control is 1.476m/s.(3)In order to realize collision avoidance of virtual coupling train formation,the anti-collision control method of the virtual coupling train based on variable parameter artificial potential field method is proposed,and simulation verification is carried out.The results show that the anti-collision control method of the virtual coupling train based on variable parameter artificial potential field method can always keep the distance of the trains in the virtual coupling train formation longer than 0.1m,which realizes the collision avoidance of the virtual train formation.(4)In order to improve the control effect of virtual coupling train,and meet the control precision of precise entry and stop in the station,the fusion method of variable parameter artificial potential field and generalized model predictive control based on characteristics and weight is proposed,and simulation verification is performed.The simulation results show that the fusion method based on weight can realize the cooperative control and collision avoidance control of virtual train formation simultaneously.The maximum speed tracking error is 1.677m/s.The average speed tracking error is 0.0062m/s.Distance tracking is always kept above 0.1m.Besides,this method solves the problem of the cooperative control of the virtual coupling train and the collision avoidance of the train in the formation,and realizes the precise stop of the virtual coupling train.There are 53 figures,6 tables and 83 references. |