Railway dynamic traffic flow organization optimization method based on arrival time limit

Abstract

The invention discloses a railway dynamic traffic flow organization optimization method based on arrival time limit. Train marshalling plans and train working diagrams are important basic technical files for guiding railway cargo transportation organizations. The method comprises the following steps of 1, inputting basic data; 2, constructing a discrete space-time network; 3, constructing a railway dynamic traffic flow organization optimization model based on the arrival time limit; and 4, generating a dynamic traffic flow organization optimization scheme and a train basic working diagram optimization scheme. According to the invention, the station daily shift plan can be automatically compiled in overall consideration, the station working efficiency is improved, and the overall transportation quality of a road network is improved; organizing and planning of whole-process transportation of the cargoes are achieved, and the cargoes can be delivered within the specified transportation time limit.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent transportation, and in particular relates to a railway dynamic vehicle flow organization optimization method based on delivery time limit. Background Art

[0002] Railway transportation has the advantages of being green, safe, economical, and all-weather, and plays a vital role in my country's social and economic development. As the current logistics market puts forward higher requirements for the punctuality and efficiency of freight transportation, it brings huge opportunities and challenges to railway transportation. The train marshaling plan and train operation diagram are important basic technical documents guiding the organization of railway freight transportation, which determine the overall transportation efficiency and transportation cost of the road network. At present, my country's railway freight transportation adopts an organizational model, and there are mainly the following problems: First, the train marshaling plan and train operation diagram are only a static traffic organization plan, which is difficult to cope with daily traffic fluctuations. Second, the daily shift plan is mainly manually compiled by the dispatchers of each station, and is not compiled from the perspective of the National Railway Group and the railway bureau as a whole, resulting in a large randomness in the traffic organization, and it is difficult to guarantee the time limit for the delivery of goods, and it is difficult to achieve system optimization. Third, for each batch of goods, the whole process of transportation organization is not planned, which makes it difficult to estimate its delivery time. Summary of the invention

[0003] In order to make up for the deficiencies of the prior art, the present invention provides a railway dynamic vehicle flow organization optimization method based on delivery time limit to solve the problems existing in the background technology.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] A railway dynamic vehicle flow organization optimization method based on delivery time limit includes the following steps:

[0006] Step 1: Basic data input;

[0007] Step 2: Construct discrete space-time network;

[0008] Step 3: Construct a railway dynamic vehicle flow organization optimization model based on the delivery time limit;

[0009] Step 4: Generate dynamic traffic organization optimization plan and train basic operation diagram optimization plan.

[0010] Furthermore, step one is specifically as follows: inputting information such as freight demand, basic road network, basic train operation diagram, station unpacking capacity, number of train sets, number of alternative operation lines as basic data, wherein the freight demand includes the departure and arrival stations, estimated departure time and the latest arrival time required by the cargo owner.

[0011] Furthermore, step 2 is specifically: expand the nodes in the basic road network on the discrete time axis to form a two-dimensional network that can represent the time-space relationship. The specific steps of constructing a discrete time-space network are as follows:

[0012] Step 1: Generate discrete space-time network nodes. There are two types of space-time networks: real nodes and virtual nodes. The real node (i, t) is extended from the station in the road network in discrete time, indicating that the train can actually depart and arrive from here; there are two types of virtual nodes, one is the node (i', t) virtualized from the real node (i, t), which is used to represent the adaptation and transfer process of the vehicle flow within the technical station i, and the other is the node (i', t) virtualized from the real node (i, t), which is used to represent the latest arrival time D of the cargo flow a. a The virtual node (i, D a ) is used to represent the delivery of cargo within the delivery deadline. Finally, n((T / K)+1) nodes are generated according to the time limit T, time scale K, and number of stations n of the discrete space-time network.

[0013] Step 2: Generate cargo flow demand. The starting node and the final node of cargo flow a are o(a)∈V and d(a)∈V respectively, and the latest delivery deadline is D a .

[0014] Step 3: Generate a travel arc (i, j, t, s)∈E, which represents the interval travel process of a train departing from station i at time t and arriving at station j at time s.

[0015] Step 4: Generate a transit arc (i,i',t,s)∈E C , represents the process of traffic flow reconfiguration and transfer within technical station i, and the transfer fee is the minimum reconfiguration and connection time θ of the traffic flow at station i i , so as to ensure that the dynamic flow distribution of the technical station meets the requirements of the reshuffling operation. For the non-reshuffling transfer train at technical station i, it is virtualized as passing through the station without stopping, and the time to arrive at other stations remains unchanged. This can avoid the reshuffling of other traffic into the train when the train in the model performs non-reshuffling technical operations at technical station i.

[0016] Step 5: Generate virtual arc (i,i,t,D a )∈E C , used to indicate the latest arrival time of cargo flow a D a The delivery within is large enough and belongs to the auxiliary arc.

[0017] Step 6: Generate waiting arc (i,i,t,s)∈E C , which is used to indicate that cargo flow a is detained at station i from time t to time s, and the waiting cost is the time scale K.

[0018] Step 7: Generate super arc (i, j, t, s) ∈ E C, directly pointing from the origin node o(a) of cargo flow a to the destination node d(a), the capacity is large enough, and the cost is a sufficiently large positive number Z to prevent the situation where a feasible solution cannot be obtained when the road network transportation capacity is insufficient. It is an auxiliary arc.

[0019] Furthermore, step three is specifically as follows: based on the aforementioned basic data and space-time network, a function model based on minimizing the overall transportation cost of the road network is constructed, and the operating line is used as a decision variable to constrain the train traction constant, the minimum number of train formations, the number of alternative operating lines, node flow balance, technical station disintegration, and formation capacity.

[0020] Prioritize the quality of freight service, so the minimum total transportation time of all vehicles in the road network is used as the objective function. The mathematical expression is:

[0021]

[0022] Where: C i,j,t,s is the transportation cost of arc (i, j, t, s) in the space-time service network, h; is a 0-1 decision variable. If cargo flow a is assigned to the arc (i, j, t, s) in the space-time network, then H a is the number of trucks required in cargo flow a, trucks.

[0023] (1) Train traction constant constraint

[0024] Considering the train transport capacity constraint as the maximum number of trains, the mathematical expression is:

[0025]

[0026] Where: y i,j,t,s is a 0-1 decision variable. If the alternative arc (i, j, t, s) in the spatiotemporal network is selected, then y i,j,t,s =1; Cap i,j,t,s is the capacity of the travel arc (i, j, t, s) in the space-time service network.

[0027] (2) Constraints on the minimum number of cars in a train formation

[0028] The train can be operated with less axles to a certain extent, but the minimum number of vehicles in the formation must be met (the detaching train and the small operation train are not affected by this constraint due to their special operation nature). The mathematical expression is:

[0029]

[0030] Where: G is the minimum number of cars in the train formation.

[0031] (3) Dynamic optimization constraints of the basic train operation diagram

[0032] In the basic train operation diagram, the transport capacity needs to have a certain reserve, so there is a limit on the number of operation lines available for alternative adjustment. The mathematical expression is:

[0033]

[0034] Where: B represents the number of running lines available for adjustment. i,j,t,s =1 means selecting this running line.

[0035] (4) Node flow balance constraints

[0036] In order to ensure that the cargo flow only chooses one transportation path, a flow balance constraint is constructed. The mathematical expression is:

[0037]

[0038] (5) Technical station dismantling capacity constraints

[0039] The mathematical expression of the dismantling capacity constraint of the medium technology station is:

[0040]

[0041] Where: N i is the number of trucks that need to arrive at station i; is the disintegration capability of technical station i within the time limit T, car.

[0042] (6) Technical station marshaling capacity constraints

[0043] The middle technical station needs to marshal the trains that depart from the station and the trains that transfer and marshal at the station. The mathematical expression of the marshaling capacity constraint is:

[0044]

[0045] Where: M i is the number of trucks that have departure demand at station i, vehicles; is the marshaling capacity of technical station i within the time limit T, car.

[0046] (7) Decision variable logical constraints

[0047] The decision variables in are all 0-1 variables, and their logical constraint expressions are

[0048]

[0049] y i,j,t,s ={0,1}⑼

[0050] Considering that the shunting locomotives of the technical station can be arranged flexibly, we combine equations (6) and (7) to simplify the shunting capacity of the technical station, and we can get

[0051]

[0052] Formula (10) can be described as the transfer and adaptation capacity of the technical station. Assuming that the shunting locomotive is equally busy and idle in each time period within the time limit T, T is 24h, then

[0053]

[0054] Where: γ i,i',t,s is the technical station i at time θ i The "transfer adaptation" capability within the car; θ i The minimum adaptation connection time of traffic flow, h.

[0055] Therefore, formula (10) can be simplified as

[0056]

[0057] Formula (12) can be described as the capacity constraint of the transit arc in the discrete space-time network.

[0058] Comparing equation (12) and equation (2), we can find that both constraints describe the capacity constraints of arcs in the space-time network. Therefore, equation (3-12) can be further processed as

[0059]

[0060] The transit arc is the product of virtualizing the technology station into two "technology stations" in the space-time network, which represents the transit adaptation process of the cargo flow within the technology station. The transportation cost is θ i , the transport capacity is γ i,i',t,s , which is essentially the same as the travel arc, so equation (13) can be incorporated into equation (1), and Cap i,j,t,s The meaning of includes not only the capacity of the travel arc, but also the capacity of the transfer arc. The model M can be further simplified into equations (1), (2), (3), (4), (5), (8), and (9).

[0061] Furthermore, step four is specifically as follows: based on the aforementioned model, a train line combination plan for freight can be obtained, and a dynamic traffic organization optimization plan and a train basic operation diagram optimization plan can be generated.

[0062] Beneficial effects of the present invention:

[0063] The present invention can realize the automatic compilation of the station's daily shift plan from an overall perspective, improve the station's work efficiency, and improve the overall transportation quality of the road network; secondly, it can realize dynamic optimization of the basic train operation diagram in combination with dynamic vehicle flow, so that it can better adapt to market demand; thirdly, it can realize the organizational planning of the entire process of cargo transportation, so that it can be delivered within the specified delivery time limit. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is an operation flow chart of the present invention. DETAILED DESCRIPTION

[0065] The present invention is described in detail below in conjunction with specific implementation modes.

[0066] Step 1: Basic data input

[0067] Specifically, information such as freight demand, basic road network, basic train operation diagram, station unpacking capacity, number of train sets, number of alternative operation lines, etc. is input as basic data. The freight demand includes the departure and arrival stations, estimated departure time and the latest arrival time required by the cargo owner.

[0068] Step 2: Constructing a discrete space-time network

[0069] Specifically, the nodes in the basic road network are expanded on the discrete time axis to form a two-dimensional network that can represent the relationship between time and space.

[0070] The nodes and arcs in discrete space-time networks have great plasticity. By giving them different meanings, they can represent more space-time states of dynamic flows, thereby simplifying the mathematical model and reducing the difficulty of solving the model.

[0071] A is a set of cargo flows, a is a batch of cargo flows transported between OD pairs, a∈A, W is a set of OD pairs, w is an OD pair, w∈W, N is a set of stations in the road network, N' is a set of virtual stations corresponding to the stations, L is a set of section lines in the road network, V is a set of nodes in the space-time service network, E is a set of arcs in the space-time service network, E C is the set of existing arcs in the space-time network, E B is a set of alternative arcs in the space-time network, T is the time limit, K is the time scale in the space-time network, i,j are stations in the network, i,j∈N, i',j' are virtual stations corresponding to i,j stations, i',j'∈N', (i,j) are arcs between adjacent points, representing the section line in the network, (i,j)∈L, (i,i') represents the transfer inside technical station i, (i,i')∈L, t,s is the moment in the space-time network, and (i,j,t,s) represents the arc in the space-time network.

[0072] The specific steps to construct a discrete space-time network are as follows:

[0073] Step 1: Generate discrete space-time network nodes. There are two types of space-time networks: real nodes and virtual nodes. The real node (i, t) is extended from the station in the road network in discrete time, indicating that the train can actually depart and arrive from here; there are two types of virtual nodes, one is the node (i', t) virtualized from the real node (i, t), which is used to represent the adaptation and transfer process of the vehicle flow within the technical station i, and the other is the node (i', t) virtualized from the real node (i, t), which is used to represent the latest arrival time D of the cargo flow a. a The virtual node (i, D a ) is used to represent the delivery of cargo within the delivery deadline. Finally, n((T / K)+1) nodes are generated according to the time limit T, time scale K, and number of stations n of the discrete space-time network.

[0074] Step 2: Generate cargo flow demand. The starting node and the final node of cargo flow a are o(a)∈V and d(a)∈V respectively, and the latest delivery deadline is D a .

[0075] Step 3: Generate a travel arc (i, j, t, s)∈E, which represents the interval travel process of a train departing from station i at time t and arriving at station j at time s.

[0076] Step 4: Generate a transit arc (i,i',t,s)∈E C , represents the process of traffic flow reconfiguration and transfer within technical station i, and the transfer fee is the minimum reconfiguration and connection time θ of the traffic flow at station i i , so as to ensure that the dynamic flow distribution of the technical station meets the requirements of the reshuffling operation. For the non-reshuffling transfer train at technical station i, it is virtualized as passing through the station without stopping, and the time to arrive at other stations remains unchanged. This can avoid the reshuffling of other traffic into the train when the train in the model performs non-reshuffling technical operations at technical station i.

[0077] Step 5: Generate virtual arc (i,i,t,D a )∈E C , used to indicate the latest arrival time of cargo flow a D a The delivery within is large enough and belongs to the auxiliary arc.

[0078] Step 6: Generate waiting arc (i,i,t,s)∈E C , which is used to indicate that cargo flow a is detained at station i from time t to time s, and the waiting cost is the time scale K.

[0079] Step 7: Generate super arc (i, j, t, s) ∈ E C , directly pointing from the origin node o(a) of cargo flow a to the destination node d(a), the capacity is large enough, and the cost is a sufficiently large positive number Z to prevent the situation where a feasible solution cannot be obtained when the road network transportation capacity is insufficient. It is an auxiliary arc.

[0080] Step 3: Construct a railway dynamic traffic organization optimization model based on delivery time limit

[0081] Specifically, based on the aforementioned basic data and space-time network, a function model based on the minimum overall transportation cost of the road network is constructed, and the operating line is used as a decision variable to constrain the train traction constant, the minimum number of train vehicles in the train formation, the number of alternative operating lines, node flow balance, technical station disintegration, and marshalling capacity.

[0082] Prioritize the quality of freight service, so the minimum total transportation time of all vehicles in the road network is used as the objective function. The mathematical expression is:

[0083]

[0084] Where: C i,j,t,s is the transportation cost of arc (i, j, t, s) in the space-time service network, h; is a 0-1 decision variable. If cargo flow a is assigned to the arc (i, j, t, s) in the space-time network, then H a is the number of trucks required in cargo flow a, trucks.

[0085] (1) Train traction constant constraint

[0086] Considering the train transport capacity constraint as the maximum number of trains, the mathematical expression is:

[0087]

[0088] Where: y i,j,t,s is a 0-1 decision variable. If the alternative arc (i, j, t, s) in the spatiotemporal network is selected, then y i,j,t,s =1; Cap i,j,t,s is the capacity of the travel arc (i, j, t, s) in the space-time service network.

[0089] (2) Constraints on the minimum number of cars in a train formation

[0090] The train can be operated with less axles to a certain extent, but the minimum number of vehicles in the formation must be met (the detaching train and the small operation train are not affected by this constraint due to their special operation nature). The mathematical expression is:

[0091]

[0092] Where: G is the minimum number of cars in the train formation.

[0093] (3) Dynamic optimization constraints of the basic train operation diagram

[0094] In the basic train operation diagram, the transport capacity needs to have a certain reserve, so there is a limit on the number of operation lines available for alternative adjustment. The mathematical expression is:

[0095]

[0096] Where: B represents the number of running lines available for adjustment. i,j,t,s =1 means selecting this running line.

[0097] (4) Node flow balance constraints

[0098] In order to ensure that the cargo flow only chooses one transportation path, a flow balance constraint is constructed. The mathematical expression is:

[0099]

[0100] (5) Technical station dismantling capacity constraints

[0101] The mathematical expression of the dismantling capacity constraint of the medium technology station is:

[0102]

[0103] Where: N i is the number of trucks that need to arrive at station i; is the disintegration capability of technical station i within the time limit T, car.

[0104] (6) Technical station marshaling capacity constraints

[0105] The middle technical station needs to marshal the trains that depart from the station and the trains that transfer and marshal at the station. The mathematical expression of the marshaling capacity constraint is:

[0106]

[0107] Where: M i is the number of trucks that have departure demand at station i, vehicles; is the marshaling capacity of technical station i within the time limit T, car.

[0108] (7) Decision variable logical constraints

[0109] The decision variables in are all 0-1 variables, and their logical constraint expressions are

[0110]

[0111] y i,j,t,s ={0,1}⑼

[0112] Considering that the shunting locomotives of the technical station can be arranged flexibly, we combine equations (6) and (7) to simplify the shunting capacity of the technical station, and we can get

[0113]

[0114] Formula (10) can be described as the transfer and adaptation capacity of the technical station. Assuming that the shunting locomotive is equally busy and idle in each time period within the time limit T, T is 24h, then

[0115]

[0116] Where: γ i,i',t,s is the technical station i at time θ i The "transfer adaptation" capability within the car; θ i The minimum adaptation connection time of traffic flow, h.

[0117] Therefore, formula (10) can be simplified as

[0118]

[0119] Formula (12) can be described as the capacity constraint of the transit arc in the discrete space-time network.

[0120] Comparing equation (12) and equation (2), we can find that both constraints describe the capacity constraints of arcs in the space-time network. Therefore, equation (3-12) can be further processed as

[0121]

[0122] In fact, the transit arc is the product of virtualizing the technology station into two "technology stations" in the space-time network, which represents the transit adaptation process of the cargo flow within the technology station. The transportation cost is θ i , the transport capacity is γ i,i',t,s , which is essentially the same as the travel arc, so equation (13) can be incorporated into equation (1), and Cap i,j,t,s The meaning of includes not only the capacity of the travel arc, but also the capacity of the transfer arc. The model M can be further simplified into equations (1), (2), (3), (4), (5), (8), and (9).

[0123] Step 4: Generate dynamic traffic organization optimization plan and train basic operation diagram optimization plan

[0124] Specifically, based on the aforementioned model, a combination plan of freight train lines can be obtained, that is, a dynamic traffic organization optimization plan and a basic train operation diagram optimization plan can be generated.

[0125] The content of the present invention is not limited to the embodiments listed, and any equivalent changes made to the technical solution of the present invention by ordinary technicians in this field after reading the description of the present invention are covered by the claims of the present invention.

Claims

1. A railway dynamic vehicle flow organization optimization method based on delivery time limit, characterized by: The steps include: Step 1: Basic data input; Step 2: Construct discrete space-time network; Step 3: Construct a railway dynamic vehicle flow organization optimization model based on the delivery time limit; Step 4: Generate dynamic traffic organization optimization plan and train basic operation diagram optimization plan.

2. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 1 is characterized by: Step one is specifically as follows: freight demand, basic road network, basic train operation diagram, station unpacking capacity, number of train sets, and number of alternative operation lines are input as basic data. The freight demand includes the departure and arrival stations, estimated departure time, and the latest arrival time required by the cargo owner.

3. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 2 is characterized in that: Step 2 is as follows: The nodes in the basic road network are expanded on the discrete time axis to form a two-dimensional network that can represent the relationship between time and space.

4. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 3 is characterized in that: Step three is as follows: Based on the aforementioned basic data and space-time network, a function model based on minimizing the overall transportation cost of the road network is constructed, and the operating line is used as a decision variable to constrain the train traction constant, the minimum number of train vehicles in the train formation, the number of alternative operating lines, node flow balance, technical station disintegration, and formation capacity.

5. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 4 is characterized in that: Step 4 is as follows: Based on the above model, the combined plan of freight train operation lines can be obtained, and the dynamic traffic organization optimization plan and the basic train operation diagram optimization plan can be generated.

6. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 5 is characterized in that: The specific steps to construct a discrete space-time network are as follows: Step 1: Generate discrete space-time network nodes. The constructed space-time network has two types: actual nodes and virtual nodes. The actual node (i, t) is extended from the station in the road network in discrete time, indicating that the train can actually depart and arrive from it. There are two types of virtual nodes. One is the node (i', t) virtualized from the actual node (i, t), which is used to represent the adaptation and transfer process of the vehicle flow within the technical station i, and the other is the node (i', t) virtualized from the actual node (i, t), which is used to represent the latest arrival time D of the cargo flow a. a The virtual node (i, D a ) is used to represent the delivery of cargo within the delivery deadline; finally, n((T / K)+1) nodes are generated according to the time limit T, time scale K, and number of stations n of the discrete space-time network; Step 2: Generate cargo flow demand. The starting node and the final node of cargo flow a are o(a)∈V and d(a)∈V respectively, and the latest delivery deadline is D a ; Step 3: Generate a travel arc (i, j, t, s) ∈ E, which represents the interval travel process of a train departing from station i at time t and arriving at station j at time s; Step 4: Generate a transit arc (i,i',t,s)∈E C , represents the process of traffic flow reconfiguration and transfer within technical station i, and the transfer fee is the minimum reconfiguration and connection time θ of the traffic flow at station i i ; Step 5: Generate virtual arc (i,i,t,D a )∈E C , used to indicate the latest arrival time of cargo flow a D a Delivery within Step 6: Generate waiting arc (i,i,t,s)∈E C , which is used to indicate that cargo flow a is detained at station i from time t to time s, and the waiting cost is the time scale K; Step 7: Generate super arc (i, j, t, s) ∈ E C , directly pointing from the origin node o(a) of cargo flow a to the destination node d(a).

7. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 6 is characterized by: There are two types of nodes in the constructed space-time network: real nodes and virtual nodes.

8. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 7 is characterized in that: In step 3, the quality of freight service is given priority, so the minimum total transportation time of all vehicles in the road network is used as the objective function. The mathematical expression is: Where: C i,j,t,s is the transportation cost of arc (i, j, t, s) in the space-time service network, h; is a 0-1 decision variable. If cargo flow a is assigned to the arc (i, j, t, s) in the space-time network, then H a is the number of trucks required in cargo flow a, cars; (1) Train traction constant constraint Considering the train transport capacity constraint as the maximum number of trains, the mathematical expression is: Where: y i,j,t,s is a 0-1 decision variable. If the alternative arc (i, j, t, s) in the spatiotemporal network is selected, then y i,j,t,s =1; Cap i,j,t,s is the capacity of the travel arc (i, j, t, s) in the space-time service network, the car; (2) Constraints on the minimum number of cars in a train formation The train can be operated with less axles to a certain extent, but the minimum number of vehicles in the formation must be met (the detaching train and the small operation train are not affected by this constraint due to their special operation nature). The mathematical expression is: Where: G is the minimum number of cars in a train formation, cars; (3) Dynamic optimization constraints of the basic train operation diagram In the basic train operation diagram, the transport capacity needs to have a certain reserve, so there is a limit on the number of operation lines available for alternative adjustment. The mathematical expression is: Where: B represents the number of running lines available for adjustment. i,j,t,s =1 means selecting the running line; (4) Node flow balance constraints In order to ensure that the cargo flow only chooses one transportation route, a flow balance constraint is constructed. The mathematical expression is: (5) Technical station dismantling capacity constraints The mathematical expression of the disintegration capacity constraint of the medium technology station is: Where: N i is the number of trucks that need to arrive at station i; is the disintegration capability of technical station i within the time limit T, car; (6) Technical station marshaling capacity constraints The medium technical station needs to marshal the station's self-organized departure trains and the trains transferred and reorganized at the station. The mathematical expression of the marshaling capacity constraint is: Where: M i is the number of trucks that have departure demand at station i, vehicles; is the marshaling capacity of technical station i within the time limit T, cars; (7) Decision variable logical constraints The decision variables in are all 0-1 variables, and their logical constraint expressions are: y i,j,t,s ＝{0，1}⑼ Considering that the shunting locomotives of the technical station can be arranged flexibly, equations (6) and (7) are combined and simplified to consider the unshunting capacity of the technical station, and the following is obtained:

9. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 8 is characterized in that: In step 3, formula (10) can be described as the transfer and adaptation capacity of the technical station. Assuming that the shunting locomotive is equally busy and idle in each time period within the time limit T, T is 24h, then: Where: γ i,i',t,s is the technical station i at time θ i The "transfer adaptation" capability within the car; θ i The minimum adaptation connection time of traffic flow, h.

10. The method for optimizing railway dynamic vehicle flow organization based on delivery time limit according to claim 9 is characterized in that: In step 3, formula (10) can be simplified as: Formula (12) can be described as the capacity constraint of the transit arc in the discrete space-time network; Comparing equation (12) and equation (2), we can find that both constraints describe the capacity constraints of arcs in the space-time network. Therefore, equation (3-12) can be further processed as: The transit arc is the product of virtualizing the technology station into two "technology stations" in the space-time network, which represents the transit adaptation process of the cargo flow within the technology station. The transportation cost is θ i , the transport capacity is γ i,i',t,s , which is essentially the same as the travel arc, so equation (13) can be incorporated into equation (1), and Cap i,j,t,s The meaning of includes not only the capacity of the travel arc, but also the capacity of the transfer arc. The model M can be further simplified into equations (1), (2), (3), (4), (5), (8), and (9).

Images