Balanced allocation method for transportation resources of Chinese and European trains

Abstract

The invention belongs to the technical field of cross-border railway cargo transportation, and particularly relates to a Chinese and European train transportation resource balanced allocation method, which comprises the following steps: acquiring cargo flow demands of Chinese and European trade cold and fresh cargos, common cargos and special cargos in a planning period, and matching different types and quantities of containers for each cargo flow; the method comprises the following steps: acquiring a basic running chart and corresponding running line use cost among stations on a channel of a Chinese and European train, use cost of different vehicle bottoms and containers, late punishment and container throwing punishment of the containers, and the lowest load factor of each station; constructing a constraint model of operation line selection, vehicle bottom application, cargo flow distribution, container allocation and decision variable range; establishing a dual objective function of operation cost and service quality; and solving the dual objective function by using an epsilon constraint method, a linearization technology and a CPLEX solver to obtain a freight scheme. According to the method, a freight scheme covering cargo flow-container flow-traffic flow-train flow can be obtained, and the waiting problem caused by insufficient transportation resources of the station can be relieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of cross-border railway freight transportation, and particularly relates to a method for balancing and deploying transportation resources of a China-Europe block train. BACKGROUND

[0002] As a core land logistics channel of the Eurasian continent, the China-Europe block train has run more than 100,000 trains and transported more than 50,000 types of goods. The transportation efficiency and resource balance of the China-Europe block train highly depend on the multi-dimensional coordination of the locomotive and car resource dynamic scheduling and distribution, the train route planning and timetable optimization, the freight allocation scheme, and the container allocation plan. The locomotive and car resource dynamic scheduling and distribution are responsible for the dynamic scheduling and distribution of locomotive and car resources. The train route planning and timetable optimization are required to determine the train route and node time based on the comprehensive path planning and timetable optimization. The freight allocation scheme is matched with the block train capacity according to the freight attributes and transportation priority. The container allocation plan is responsible for the cross-period allocation of empty and full containers to ensure the loading demand and turnover efficiency. The existing scheme preparation method adopts an independent optimization mode, lacks a multi-dimensional coordination mechanism, and is prone to cause resource mismatch (such as the coexistence of idle car bottom and container shortage), transportation cost redundancy, and service target conflict.

[0003] Among the existing China-Europe block train related invention patents, a China-Europe block train railway logistics data full-process tracing method and system (CN202410921166.0) and a China-Europe block train operation statistics method and system (CN202011401581.1) provide a data statistics method. A China-Europe block train transportation network construction method considering hub node failure (CN202010156928.4) and a China-Europe block train domestic collection and transportation system optimization method based on double-layer planning (CN202010179829.8) provide a freight transportation network construction method. A China-Europe block train empty container allocation optimization method based on container sharing (CN202010157230.4) provides an empty container allocation method, but there is no China-Europe block train freight transportation plan preparation method.

[0004] Among the existing railway freight related invention patents, a railway freight real-time comprehensive state calculation method and system (CN202210576250.4) and a railway freight service dynamic matching method and system based on customer characteristics (CN202310486661.9) provide a freight service investigation method. A container collaborative allocation method based on idle container intermodal transportation (CN202410906122.0) and a model and method for railway freight empty car allocation for random demand (CN202210317640.X) provide a resource scheduling method. A port and railway freight collaborative scheduling management system (CN202311347753.5) and a railway freight train operation scheme optimization method and device under uneven transport capacity conditions (CN202211686896.4) provide a freight train scheduling method.

[0005] The current Central and Eastern Europe block train cross-border railway transportation system has the following technical defects: firstly, the existing railway resource scheduling technology cannot dynamically adapt to the differentiated storage and transportation conditions (temperature control, shock prevention) and transportation time limit (emergency order, regular block train) of high value-added goods (such as cold chain medicine) and bulk materials (such as grain), resulting in low utilization rate of car bottom and container and unstable transportation time limit; secondly, there is a lack of collaborative optimization mechanism between car bottom operation plan, running line selection scheme, freight flow allocation scheme and container allocation plan, resulting in imbalance between arrival and departure time sequence of transportation resources at hub stations or port stations, coexistence of local redundancy and shortage of resources; thirdly, the existing scheme preparation logic does not solve the problem of collaborative optimization of double targets of transportation cost control and service quality improvement, and it is difficult to meet the differentiated needs of block train operators and consignors. SUMMARY

[0006] Therefore, the purpose of the present application is to provide a Central and Eastern Europe block train transportation resource balanced allocation method, which synchronously prepares car bottom operation plan, running line selection scheme, freight flow allocation scheme and container allocation plan, and takes operation cost and service quality as double targets to obtain a double-target balanced freight transportation scheme. The freight transportation scheme can provide a Central and Eastern Europe block train freight transportation scheme covering freight flow-container flow-car flow-train flow, can avoid mismatch between arrival and departure number of car bottom and container resources at border stations and hub stations, and can significantly alleviate the waiting problem caused by insufficient transportation resources at stations.

[0007] The present application provides a Central and Eastern Europe block train transportation resource balanced allocation method, comprising:

[0008] S1, obtaining the freight flow demand of Central and Eastern Europe trade cold fresh goods, ordinary goods and special goods in the planning period, and matching different types and quantities of containers for each freight flow;

[0009] S2, obtaining the basic running diagram between each station on the Central and Eastern Europe block train channel and the corresponding running line use cost, the use cost of different car bottom and container, the late arrival penalty and container rejection penalty of container, and the minimum full load rate of each station;

[0010] S3, constructing a constraint model of running line selection, car bottom operation, freight flow allocation, container allocation and decision variable range based on the data obtained in S1;

[0011] S4, establishing a double target function of operation cost and service quality based on the data obtained in S2;

[0012] S5, using the constraint model of S3, using the epsilon constraint method, linearization technology and CPLEX solver to solve the double target function of S4 to obtain a freight transportation scheme.

[0013] Furthermore, in step S3, based on the data obtained in S1, a constraint model for operation line selection, vehicle bottom utilization, cargo flow allocation, container allocation, and decision variable range is constructed, including:

[0014] S31. By properly attaching the traffic flow to the running line, the model running line selection constraint is:

[0015] y a ≤δ a ,

[0016] Where A represents the set of running lines of all sections on the line, δ a =1 means that the running line a∈A is occupied, otherwise it is 0, y a =1 means that the running line a∈A has been selected, otherwise it is 0;

[0017] S32. By limiting the number of arriving vehicles at each station to be consistent with the number of departing vehicles, the number of vehicles in the operating section is always balanced, and the vehicle utilization constraint is modeled as:

[0018]

[0019] Where x al =1 indicates running line has been selected, otherwise 0;

[0020] S33. By limiting the inflow of different types of containers corresponding to different cargo flows to be equal to the outflow, the cargo flow allocation constraint is modeled as:

[0021]

[0022] Where A - ,A + They represent the set of departure and arrival lines of the station, f kua represents the flow size of container types u∈U in cargo flow k∈K allocated to running line a∈A, where set K represents cargo flow demand and set U represents container types;

[0023] S34. By limiting the outflow of different types of containers at any station to be equal to the inflow, the modeled container allocation constraint is:

[0024]

[0025] Where q ua represents the number of containers of type u∈U on the running line a∈A;

[0026] S35, the decision variable range defines the 0 / 1 decision variables for the selection of the running line and the use of the vehicle bottom:

[0027] y a ,x al ∈{0,1},

[0028] Another aspect limits the number of flow allocations and container allocations:

[0029] f kua ∈[0,D ku ],

[0030] q ua ∈[0,C ua ],

[0031] where D ku denotes the number of containers of type u∈U required by flow k∈K, and C ua denotes the carrying capacity of line a∈A for containers of type u∈U.

[0032] Further, the step S4 of establishing a dual objective function of operation cost and service quality based on the data obtained in S2 comprises:

[0033] S41, based on the operation line usage cost, the car bottom usage cost and the container usage cost, an operation cost objective function is established:

[0034]

[0035] where c al denotes the cost of selecting the connecting arc segment a, l∈A for the car bottom, c denotes the usage cost of a single car bottom, w ua denotes the cost of selecting the operation line a∈A for a single container of type u∈U;

[0036] S42, based on the container late arrival penalty cost, a service quality objective function is modeled:

[0037]

[0038] where t denotes the late arrival time of the destination of flow k∈K, denotes the train arrival time of the operation line a∈A, T k denotes the required arrival time of flow k∈K, and p ku denotes the unit late arrival penalty of flow k∈K type u∈U.

[0039] Further, the step S5 of solving the dual objective function of S4 by using the ε constraint method, linearization technique and CPLEX solver comprises:

[0040] S51, add the service quality target as a constraint by using the ε constraint method, so as to convert the double target function into a single target function of operation cost;

[0041] S52, convert the nonlinear function in the service quality target function into the following integer linear programming model by using the large M method of linearization technology:

[0042]

[0043] δ ≥0

[0044]

[0045] δ≤Mb

[0046] b∈{0,1};

[0047] In the formula, the decision variable δ is used to replace the nonlinear function, and M represents a maximum positive integer;

[0048] S53, after mixing the single target function of operation cost and the integer linear programming model, the CPLEX solver is used for solving to obtain a freight scheme.

[0049] Further, the step S53 of using the CPLEX solver for solving to obtain the freight scheme comprises:

[0050] The mixed model of the single target function of operation cost and the integer linear programming model is solved by using the CPLEX solver to obtain each decision variable;

[0051] The decision variable x al ∈{0,1}, The decision variable y a {0,1}, The decision variable f kua , The decision variable q ua , The decision variable q ua ,

[0052] The method for balancing and deploying transportation resources of the China-Europe block train according to the application can obtain a double-target balanced freight scheme by synchronously preparing a car bottom use plan, a running line selection scheme, a freight flow allocation scheme and a container deployment plan, and taking operation cost and service quality as double targets. The freight scheme can provide a China-Europe block train freight transportation scheme covering freight flow-container flow-car flow-train flow, can avoid the mismatch between the number of car bottom and container resources arriving and leaving the border station and hub station, and can significantly alleviate the waiting problem caused by insufficient transportation resources at the station. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flow chart of a method for balancing and allocating transport resources of China-Europe freight trains provided in an embodiment of the present application is shown;

[0054] Figure 2 A schematic diagram of the China-Europe freight train transport network provided by an embodiment of the present application is shown;

[0055] Figure 3 A Pareto frontier curve diagram obtained by using the ε constraint method in the prior art provided by an embodiment of the present application is shown;

[0056] Figure 4 A schematic diagram of the operating line selection and vehicle bottom turnover provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of this technical solution more clear, the following technical solution is further described in detail in conjunction with specific implementation methods. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of this technical solution.

[0058] Example 1:

[0059] Please refer to Figure 1 The flowchart of the method for balancing and allocating transport resources of China-Europe Railway Express is shown in FIG. Figure 1 As shown, the method includes:

[0060] S101. Obtain the cargo flow demand for refrigerated goods, general goods, and special goods in China-Europe trade during the planning period, and match different types and quantities of containers for each cargo flow.

[0061] S102. Obtain the basic operation diagram between stations on the China-Europe Express corridor and the corresponding operation line usage costs, usage costs of different vehicle chassis and containers, penalties for late arrival and abandoned containers, and the minimum load rate of each station.

[0062] S103. Based on the data obtained in S101, a constraint model for operation line selection, vehicle bottom utilization, cargo flow distribution, container allocation, and decision variable range is constructed.

[0063] In this step, the purpose of building the constraint model is to integrate and optimize the operating line selection plan, vehicle bottom utilization plan, cargo flow distribution plan and container allocation plan, and to achieve balanced allocation of vehicle bottom and container transportation resources.

[0064] In specific implementation, the constraint models for line selection, vehicle bottom utilization, cargo flow distribution, container allocation, and decision variable range can be constructed in the following ways:

[0065] S1031, By reasonably hanging the train flow to the running line, the running line selection constraint is modeled as:

[0066] y a ≤δ a ,

[0067] where A represents the set of running lines of all sections on the line, δ a = 1 indicates that the running line a∈A has been occupied, otherwise it is 0, y a = 1 indicates that the running line a∈A has been selected, otherwise it is 0.

[0068] S1032, By limiting the number of arriving and departing train bottoms at each station to keep the number of train bottoms on the running section balanced at all times, the train bottom utilization constraint is modeled as:

[0069]

[0070] where x al = 1 indicates that the running line has been selected, otherwise it is 0.

[0071] S1033, By limiting the inflow and outflow of different types of containers corresponding to different freight flows to be equal, the freight flow distribution constraint is modeled as:

[0072]

[0073] where A - ,A + represent the set of departure and arrival running lines at the station, f kua represents the flow size of container type u∈U in freight flow k∈K allocated to running line a∈A, where set K represents freight flow demand and set U represents container type.

[0074] S1034, By limiting the outflow and inflow of different types of containers at any station to be equal, the container allocation constraint is modeled as:

[0075]

[0076] where q ua represents the number of container type u∈U on running line a∈A.

[0077] S1035, The decision variable range defines the 0 / 1 decision variables of running line selection and train bottom utilization on the one hand:

[0078] y a ,x al ∈{0,1},

[0079] The other aspect limits the number of flow distribution and container allocation:

[0080] f kua ∈[0,D ku ],

[0081] q ua ∈[0,C ua ],

[0082] where D ku denotes the number of containers of type u∈U required by flow k∈K, C ua denotes the carrying capacity of the route a∈A for containers of type u∈U.

[0083] S104, based on the data obtained in S102, establish a dual objective function of operation cost and service quality;

[0084] In specific implementation, the dual objective function of operation cost and service quality can be established in the following manner:

[0085] S1041, based on the route usage cost, the car bottom usage cost and the container usage cost, establish an operation cost objective function:

[0086]

[0087] where c al denotes the cost of selecting the connecting arc segment a, l∈A for the car bottom, c denotes the usage cost of a single car bottom, w ua denotes the cost of selecting the route a∈A for a single container of type u∈U.

[0088] S1042, based on the container late arrival penalty cost, model a service quality objective function:

[0089]

[0090] where t denotes the late arrival time of the destination of flow k∈K, t denotes the train arrival time of the route a∈A, T k denotes the required arrival time of flow k∈K, ρ ku denotes the unit late arrival penalty of flow k∈K of type u∈U.

[0091] S105, using the constraint model of S103, solve the dual objective function of S104 by using the ε constraint method, linearization technique and CPLEX solver to obtain a freight transport scheme.

[0092] In the implementation, the double-objective function can be solved by the following method:

[0093] S1051, the quality of service target is added as a constraint by using the epsilon constraint method, so as to convert the double-objective function into a single-objective function of operation cost.

[0094] S1052, the nonlinear function in the quality of service target function is converted into the following integer linear programming model by using the large M method of linearization technology:

[0095]

[0096] δ ≥0

[0097]

[0098] δ≤Mb

[0099] b∈{0,1};(9)

[0100] In the formula, the decision variable δ is used to replace the nonlinear function, and M represents a maximum positive integer.

[0101] In this step, the δ in the integer linear programming model is equivalent to the δ in the nonlinear function. The δ in the integer linear programming model is equivalent to the δ in the nonlinear function. The δ in the integer linear programming model is equivalent to the δ in the nonlinear function. The δ in the integer linear programming model is equivalent to the δ in the nonlinear function.

[0102] S1053, after mixing the single-objective function of operation cost and the integer linear programming model, the CPLEX solver is used for solving to obtain the freight scheme.

[0103] In this step, the CPLEX solver can be called by MATLAB or JAVA for efficient solving, and then the solving result is converted into the freight scheme.

[0104] In the implementation, the freight scheme can be obtained by the following method:

[0105] Step 201, the mixed model of the single-objective function of operation cost and the integer linear programming model is solved by using the CPLEX solver to obtain each decision variable.

[0106] Step 202, the decision variable x al ∈{0,1}, is used to determine the car bottom turnover use, the decision variable y a {0,1}, is used to determine the selected running line, and the decision variable f kua , Determine the container used by the freight flow and the carried route, using decision variable q ua , Determine the type and quantity of containers to obtain the freight transport scheme.

[0107] In this step, the route selection scheme, car bottom operation plan, freight flow allocation scheme and container allocation plan are obtained by solving the decision variables.

[0108] Example two,

[0109] Please refer to the schematic diagram of the Europe-China block train transport network as Figure 2 shown. The figure shows the process of freight marshalling, car bottom turnover and reloading operation. Specifically, freight type 1 and freight type 2 are marshalled at hub station A and run to border station a for reloading; after adjusting the track gauge, the train runs to hub station D and is finally disassembled and transported to different destinations 4 and 5 for different freight types.

[0110] Please refer to the Pareto frontier curve diagram obtained by using the ε constraint method in the prior art as Figure 3 shown. Each point in the figure corresponds to a Europe-China block train freight transport scheme, and the company operating cost and freight service quality (freight penalty cost) of different schemes are different. As Figure 3 shown, the increase in the company operating cost target will reduce the transport service quality target, indicating that the transport scheme cannot achieve the best cost and quality at the same time.

[0111] Therefore, using the Europe-China block train transport resource balanced allocation method of the present application, the route selection and car bottom turnover situation schematic diagram as Figure 4 shown is obtained. Figure 4 The horizontal axis represents the time period, and the vertical axis represents the stations. B-C and A-C are both marshalling sections, and the outbound route of the Europe-China block train is black and the return route is red. To meet the container transport demand, in addition to the B-C and A-C marshalling sections, two Europe-China block trains need to be run on the outbound and return routes in other sections. For example, the routes numbered 2, 12, 22, 32, 42, 52, and 62 are the hanging line schemes of the first outbound freight flow, and the routes numbered 3, 14, 23, 33, 43, 53, and 63 are the hanging line schemes of the second outbound freight flow; and the routes numbered 67, 57, 47, 37, 27, 17, and 7 are the hanging line schemes of the first return freight flow, and the routes numbered 68, 58, 48, 38, 28, 19, and 8 are the hanging line schemes of the second return freight flow. In addition, the D-L section in the figure shows the car bottom turnover situation, and the blue and yellow lines are the uplink and downlink routes of the two car bottoms. To meet the balanced use of car bottoms, the number of arriving car bottoms at D station and the number of departing car bottoms at L station are consistent.

[0112] The above merely describes the preferred embodiments of the present application, and for those skilled in the art, many changes can be made to the specific implementation and application range based on the technical content of the present application, as long as the changes do not deviate from the concept of the present application, and all belong to the protection scope of the present application.

Claims

1. A method for balancing and allocating transport resources of China-Europe freight trains, characterized in that: The method comprises: S1. Obtain the cargo flow demand for fresh goods, general cargo, and special cargo in China-Europe trade during the planning period, and match different types and quantities of containers for each cargo flow; S2. Obtain the basic operation diagram between each station on the China-Europe Express corridor and the corresponding operation line usage costs, usage costs of different vehicle chassis and containers, penalties for late arrival and abandoned containers, and the minimum load factor of each station; S3, based on the data obtained in S1, builds a constraint model for line selection, vehicle bottom utilization, cargo flow allocation, container allocation, and decision variable range; S4, establishes a dual objective function of operating cost and service quality based on the data obtained in S2; S5. Using the constraint model of S3, the ε constraint method, linearization technology, and CPLEX solver are used to solve the dual objective function of S4 and obtain the freight plan.

2. The method according to claim 1, wherein In step S3, based on the data obtained in step S1, a constraint model for operation line selection, vehicle bottom utilization, cargo flow allocation, container allocation, and decision variable range is constructed, including: S31. By properly attaching the traffic flow to the running line, the model running line selection constraint is: Where A represents the set of running lines of all sections on the line, δ a =1 means that the running line a∈A is occupied, otherwise it is 0, y a =1 means that the running line a∈A has been selected, otherwise it is 0; S32. By limiting the number of arriving vehicles at each station to be consistent with the number of departing vehicles, the number of vehicles in the operating section is always balanced, and the vehicle utilization constraint is modeled as: Where x al =1 indicates running line has been selected, otherwise 0; S33. By limiting the inflow of different types of containers corresponding to different cargo flows to be equal to the outflow, the cargo flow allocation constraint is modeled as: Where A - ,A + They represent the set of departure and arrival lines of the station, f kua represents the flow size of container types u∈U in cargo flow k∈K allocated to running line a∈A, where set K represents cargo flow demand and set U represents container types; S34. By limiting the outflow of different types of containers at any station to be equal to the inflow, the modeled container allocation constraint is: Where q ua represents the number of containers of type u∈U on the running line a∈A; S35, the decision variable range defines the 0 / 1 decision variables for the selection of the running line and the use of the vehicle bottom: On the other hand, it limits the number of cargo distribution and container deployment: Where D ku C represents the number of containers of type u∈U required for cargo flow k∈K, ua It represents the carrying capacity of the operating line a∈A for type u∈U containers.

3. The method according to claim 1, wherein In step S4, a dual objective function of operating cost and service quality is established based on the data obtained in step S2, including: S41. Based on the cost of using the running line, the cost of using the vehicle bottom, and the cost of using the container, establish the operating cost objective function: Where c al represents the cost of selecting the connecting arc segment a,l∈A under the vehicle bottom, c represents the cost of using a single vehicle bottom, w ua represents the cost of a single container of type u∈U choosing to run line a∈A; S42. Model the service quality objective function based on the penalty cost of container late arrival: Where, represents the destination late arrival time of cargo flow k∈K, represents the train arrival time of line a∈A, T k represents the required arrival time of cargo flow k∈K, ρ ku represents the unit late arrival penalty for cargo flow k∈K type u∈U.

4. The method according to claim 1, wherein In step S5, the dual objective function of S4 is solved by using the ε constraint method, linearization technology, and CPLEX solver, including: S51. Using the ε constraint method, the service quality objective is added as a constraint, thereby transforming the dual objective function into a single objective function of operating cost; S52. Using the Big M method of linearization technology, the nonlinear function in the service quality objective function is transformed into the following integer linear programming model: δ≥0 δ≤Mb b∈{0,1}； In the formula, the decision variable δ is used to replace the nonlinear function, and M represents a maximum positive integer; S53. After mixing the single objective function of operating cost with the integer linear programming model, the CPLEX solver is used to solve the problem to obtain the freight plan.

5. The method according to claim 4, wherein In step S53, the CPLEX solver is used to solve the problem to obtain a freight solution, including: The hybrid model of the single objective function of operating cost and the integer linear programming model is solved using the CPLEX solver to obtain the decision variables; Using decision variables Determine the vehicle bottom turnover utilization and use decision variables Determine the selected operating line using decision variables Determine the containers and routes used for cargo flow, and use decision variables Determine the type and quantity of containers to arrive at a freight solution.

Images