Heavy-load train operation strategy optimization method and system based on group control technology

Abstract

The present invention relates to the field of transportation technology, and discloses a method and system for optimizing the operation strategy of heavy-load trains based on group control technology, so as to rationally plan the operation strategy, reduce the time efficiency of heavy-load train cargo transportation, and improve the line transportation capacity. The method comprises: step S1, determining a heavy-load railway line to be optimized with the characteristic of "directly departing from a technical station"; step S2, constructing a target optimization model that minimizes the weighted sum of economic cost and total in-transit transportation time of goods based on the supply and demand relationship, arrival time constraint, skylight time, number of locomotives, and the comprehensive importance of transportation demand; step S3, solving the group train operation strategy corresponding to the Pareto optimality in the target optimization model based on the simulated annealing algorithm; wherein the number of trains in each group is the same and greater than or equal to 2.

Description

Technical Field

[0001] The present invention relates to the field of transportation technology, and in particular to a method and system for optimizing heavy-load train operation strategies based on group control technology. Background Art

[0002] The transport organization concept for ensuring the timeliness of heavy-haul rail transport is to prioritize meeting cargo arrival deadlines and rationally utilize rail transport capacity to maximize cargo delivery. Based on this new concept for railway freight transport, railways can fully transition from a "producer-centric" to a "consumer-centric" approach, adapting to the demands of economic and social development. Therefore, it is crucial to strengthen research on time-sensitive railway freight transport organization. Furthermore, the continued development of my country's heavy-haul rail lines requires further improvements in line capacity and total train volume.

[0003] Heavy-haul train group operation control technology can improve transportation efficiency, reduce operating costs, further enhance the level of intelligent train operation control, and reduce the tracking distance between trains. Therefore, heavy-haul train operation strategies based on group control technology can effectively improve freight timeliness and train transportation capacity. Summary of the Invention

[0004] The purpose of the present invention is to disclose a heavy-load train operation strategy optimization method and system based on group control technology, so as to rationally plan the operation strategy, reduce the time efficiency of heavy-load train cargo transportation, and improve the line transportation capacity.

[0005] To achieve the above objectives, the present invention discloses a heavy-load train operation strategy optimization method based on group control technology, comprising:

[0006] Step S1: Determine a dedicated heavy-haul railway line to be optimized that has the feature of "directly departing from a technical station";

[0007] Step S2: Based on the supply-demand relationship, delivery time constraints, window time, number of locomotives, and the comprehensive importance of transportation demand, a target optimization model is constructed to minimize the weighted sum of economic cost and total in-transit transportation time of the goods;

[0008] Step S3: solving the group train operation strategy corresponding to the Pareto optimality in the target optimization model based on a simulated annealing algorithm; wherein the number of trains in each group is the same and greater than or equal to 2.

[0009] Preferably, step S2 includes:

[0010] Step S21: Set the following parameters related to the target optimization model:

[0011] I is the set of network stations, i is the station number, where i∈I; A is the set of network sections, a is the section number, where a∈A; l ais the length of section a; M is the number of groups running on a single day, m is the group number, where m∈M; N is the number of unit heavy-load trains in the group, n is the train number, where n∈N; D is the transport demand set, d is the transport demand number, where d∈D; S d represents the transportation demand d actual supply, P d represents the theoretical demand for transportation demand d; V represents the segment with the starting and ending points i-1 and i respectively; normal Indicates the normal operating speed of the train; V awa y represents the train disassembly and operation speed; c represents the unit operation cost; C represents the fixed operation cost at the first station; t1 represents the departure interval between groups; t2 represents the departure interval between trains in a group; TC s Indicates the sunroof start time, TC e Indicates the end time of the skylight; H indicates the number of traction locomotives required for the unit train; H max Indicates the maximum number of traction locomotives in the road network;

[0012] represents the departure time of train n in group m, represents the arrival time of train n in group m; U m,n Indicates the load of the unit heavy-load train; w d Indicates the comprehensive importance of transportation demand d; Lx indicates the transportation distance of transportation demand d; Tm ,n represents the transit time of the entire group m, T' m,n represents the en route running time of train n of group m with dispersed arrivals;

[0013] The relevant decision variables include: It means that if station i is the terminal station of transport demand d, it is 1, otherwise it is 0; It indicates that if segment a belongs to the path ending at station i, it is 1, otherwise it is 0; It means that if the train n of group m belongs to the transport demand d, it is 1, otherwise it is 0; It means that if the transportation demand of the entire group m is d, it is 1, otherwise it is 0;

[0014] Step S22: The model optimization objective is divided into two parts: economic cost and cargo transportation time. The economic cost is divided into the departure cost at the originating technical station and the loss cost under supply and demand matching. The departure cost at the originating technical station is expressed as The loss cost under supply and demand matching is expressed as in, The total transit time of freight is composed of three parts: the total transit time of freight of the entire group of arrival type trains, the total departure interval time, and the total transit time of the dispersed arrival type trains. The specific function is expressed as:

[0015] Step S23: Set the following constraints on the target optimization model:

[0016] The supply and demand constraints of cargo volume are as follows: S d ≥P d ,in,

[0017] Delivery deadline constraints, specifically:

[0018] Route skylight time constraints, specifically:

[0019] Locomotive quantity constraints, specifically:

[0020] Preferably, step S3 includes:

[0021] (31) Initial solution generation, the generation rules are as follows:

[0022] Step 311. Randomly generate (M, N) within a specific range 2 The binary combination of the values ​​is decoded to obtain the decimal value (M, N), that is, the initial network operation group M and the number of trains in the group N are obtained;

[0023] Step 312. Generate a 0-1 variable two-dimensional matrix v for the entire group based on the initial (M, N) and randomly generate a 0-1 two-dimensional matrix τ for the dispersed arrival of the group train based on v;

[0024] Step 313. Based on the two-dimensional matrices v and τ representing the group's entire or scattered arrival information in Step 312, generate a three-dimensional matrix γ representing the specific arrival information of each train in the group, considering that the departure time of the first train in the group is the window end time TC e , combined with the group departure time and the train departure time t1 and t2 within the group, generate the group train departure time two-dimensional matrix T start , based on the relevant parameters {l a ,L d ,V normal ,V away ,d,U m,n} and decision variables Calculate the travel time T of various groups of arrival type trains and dispersed arrival type group trains respectively m and T' m,n ,in, Generate group train arrival schedule T end ;

[0025] Step 314. The above operation strategy (γ, T start ,T end ) complies with the supply and demand constraints of goods quantity S d ≥P d , delivery deadline constraints Route skylight time constraints and Locomotive quantity constraints That is, it is used as the initial solution, otherwise it returns to Step 311 to regenerate;

[0026] (32) Determine the energy function minZ = g·E+j·R, where E and R represent economic cost and transportation time, respectively, g and j represent the weights corresponding to economic cost and transportation time, respectively, and minZ represents that the optimization objective of the target optimization model is to minimize the weighted sum of economic cost and total in-transit transportation time of the goods;

[0027] (33) Generation of Neighborhood Solutions

[0028] (M,N) generated for the initial solution 2 The binary combination is probabilistically perturbed, and a new combination (M, N) is obtained by decoding, and it is judged whether the new combination meets the constraint conditions; if so, then randomly generate according to Step 312-Step 314 of (31) and output the neighborhood solution; if not, the original (M, N) is subjected to random perturbation. 2 The binary combination is probabilistically perturbed again until the constraints are fully satisfied and the neighborhood solution is then sought. The iterative solution algorithm for the neighborhood solution is as follows:

[0029] Step 331. Determine the initial temperature T0, record the current cooling times k = 0, and go to Step 332;

[0030] Step 332. Let the number of iterations be n = 0. At the initial temperature, generate the initial strategy solution X0 according to the methods in (31)-(32) above and calculate the energy function value, then go to Step 333.

[0031] Step 333. After k cooling times, at temperature T k Next, let the solution after the n-1th iteration be X n-1 , in order to obtain the neighborhood solution X n , randomly update the MN value combination and calculate the corresponding function value energy Z(X n ), go to Step 334;

[0032] Step 334. Use the Metropolis criterion to test the current strategy solution Xx. If Z(X n )-Z(Xn-1 )<0, then accept X unconditionally n Instead of X n-1 ; If Z(X n )-Z(X n-1 )>0, then with probability γ n (T k ) accepts the neighborhood solution; γ n (T k )The solution formula is:

[0033]

[0034] Go to Step 335;

[0035] Step 335. At the corresponding temperature, set the upper limit of the number of iteration steps to When the number of iterations at the temperature reaches the upper limit, go to Step 336; otherwise, go to Step 333;

[0036] Step 336. Algorithm convergence termination judgment: The set termination criterion is that the external temperature is lower than the specified threshold. If the criterion is met, the algorithm ends. Otherwise, go to Step 337;

[0037] Step 337. Cool down according to the following formula:

[0038] T k+1 =p·Tk

[0039] T k After the temperature is cooled, the temperature is updated. At this time, the number of cooling times k=k+1 is recorded, the number of iteration steps n=0 is set, and the process returns to Step 333. The loop iteration is performed at the new temperature, and finally the optimal operation strategy is obtained under the set temperature and the internal iteration number.

[0040] To achieve the above-mentioned purpose, the present invention also discloses a heavy-load train operation strategy optimization system based on group control technology, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method when executing the computer program.

[0041] The present invention has the following beneficial effects:

[0042] Compared with conventional operation strategies, the group operation strategy disclosed in the present invention can reasonably plan the operation strategy while taking various factors into consideration, better meet the timeliness requirements of goods, and transport more goods at a lower economic cost, thereby improving the economic benefits of both the railway and the cargo owners.

[0043] The present invention will be described in further detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are intended to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings:

[0045] Figure 1 This is a schematic diagram of the group operation strategy disclosed in an embodiment of the present invention.

[0046] Figure 2 It is a schematic diagram of the relationship between the dedicated heavy-haul railway line and supply and demand disclosed in an embodiment of the present invention.

[0047] Figure 3 It is a flow chart of a heavy-load train operation strategy optimization method based on group control technology disclosed in an embodiment of the present invention.

[0048] Figure 4 This is a schematic diagram of the branch point numbering of the Shuohuang line in the disclosed example of the embodiment of the present invention. DETAILED DESCRIPTION

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. However, the present invention can be implemented in many different ways as defined and covered by the claims.

[0050] Example 1

[0051] This embodiment discloses a method for optimizing heavy-load train operation strategy based on group control technology.

[0052] The heavy-load train group operation strategy disclosed in this embodiment takes into account the background that the freight transportation demand is presented as a radial type from the technical starting station to each support station, such as Figure 1 As shown, the group operation mode is based on 5,000-ton trains, and the number of trains in each group is fixed and the same. The trains run in groups at certain intervals. During the operation in the section and station, the tail car of the group automatically disassembles the train group in real time.

[0053] exist Figure 1 In the example, Q1, Q2, Q3 and Q4 are four groups with the same departure station. Each group has four trains. The four trains in groups Q1 and Q3 have the same terminal station. In group Q2, only two direct trains have the same terminal station. In group Q4, the four trains have different terminal stations. When the terminal stations of the trains in a group are not completely consistent, the trains that are disassembled at the corresponding station will run at the disassembly speed. Figure 1 The different slopes from the normal operating speed can be regarded as representing different de-encoding operating speeds.

[0054] like Figure 2As shown, this embodiment considers the matching problem of heavy-haul railway transportation supply capacity and freight demand for a heavy-haul railway dedicated line with the characteristic of "directly departing from a technical station" under the background that the freight transportation demand is radial from the technical departure station to each branch station, and fully considers the supply and demand relationship, delivery time constraints, window time, number of locomotives and other constraints, introduces the comprehensive weights of various freight demands with delivery deadline, distance between two stations, cargo demand and customer level as reference indicators, constructs a multi-objective optimization model for the railway transportation enterprise to minimize the weighted sum of fixed operation and variable loss costs and total in-transit transportation time of goods under supply and demand matching, and uses simulated annealing algorithm to solve the group train operation strategy under Pareto optimality.

[0055] Based on the above purpose, Figure 3 As shown, the heavy-load train operation strategy optimization method based on group control technology in this embodiment includes the following steps:

[0056] Step S1: Determine a dedicated heavy-load railway line to be optimized that has the feature of "directly departing from a technical station".

[0057] Step S2: Based on the supply-demand relationship, delivery time constraints, window time, number of locomotives, and the comprehensive importance of transportation demand, a target optimization model is constructed to minimize the weighted sum of economic cost and total in-transit transportation time of the goods.

[0058] Preferably, this step specifically includes:

[0059] Step S21: Set the following parameters related to the target optimization model:

[0060] I is the set of network stations, i is the station number, where i∈I; A is the set of network sections, a is the section number, where a∈A; l a is the length of section a; M is the number of groups running on a single day, m is the group number, where m∈M; N is the number of unit heavy-load trains in the group, n is the train number, where n∈N; D is the transport demand set, d is the transport demand number, where d∈D; S d represents the transportation demand d actual supply, P d represents the theoretical demand for transportation demand d; V represents the segment with the starting and ending points i-1 and i respectively; normal Indicates the normal operating speed of the train; V away represents the train disassembly and operation speed; c represents the unit operation cost; C represents the fixed operation cost of the first station; t1 represents the departure interval between groups; t2 represents the departure interval between trains in a group; TC s represents the skylight start time, TCe represents the skylight end time; H represents the number of traction locomotives required for the unit train; H max Indicates the maximum number of traction locomotives in the road network.

[0061] represents the departure time of train n in group m, represents the arrival time of train n in group m; U m,n Indicates the load of the unit heavy-load train; w d Indicates the comprehensive importance of transportation demand d; L d The transport distance representing the transport demand d; Tm ,n represents the transit time of the entire group m, T' m,n represents the en route running time of train n of group m that arrives dispersedly.

[0062] The relevant decision variables include: It means that if station i is the terminal station of transport demand d, it is 1, otherwise it is 0; It indicates that if segment a belongs to the path ending at station i, it is 1, otherwise it is 0; It means that if the train n of group m belongs to the transport demand d, it is 1, otherwise it is 0; It means that if the transportation demand of the entire group m is d, it is 1, otherwise it is 0.

[0063] Step S22: Divide the model optimization objective into two parts: economic cost and cargo transportation time. The economic cost is divided into fixed cost and variable cost. The total fixed cost is expressed as The total variable cost is expressed as in, The total transit time of freight is composed of three parts: the total transit time of freight of the entire group of arrival type trains, the total departure interval time, and the total transit time of the dispersed arrival type trains. The specific function is expressed as:

[0064] In addition, in this embodiment, the biggest difference between conventional operation and group operation is that conventional operation treats each train as a group operation, that is, the first train of the group is also the last train of the group. The difference in the solution of the model is mainly reflected in the second part of the model objective, that is, Y'=T' z +T' q .

[0065] In this step, the model optimization objectives are divided into two parts: economic cost and cargo in-transit time. The economic cost is divided into the departure cost at the departure technical station and the loss cost under supply and demand matching. Generally speaking, the operation cost is divided into two parts: fixed cost and variable cost. Considering that the problem studied in this paper is about the situation under supply and demand matching, the variable cost temporarily ignores the other components and considers that the loss cost caused by the supply and demand matching difference is generally related to the number of trains, the number of kilometers operated and the tonnage of cargo. Fixed cost refers to the fixed expenses incurred each time materials are gathered, loaded, assembled and dispatched, which do not change with the amount of work, such as machinery and equipment costs, water and electricity costs, information equipment costs, etc., and it is assumed that these fixed costs remain unchanged for each operation, that is, the departure cost at the departure technical station considered in this invention.

[0066] Step S23: Set the following constraints on the target optimization model:

[0067] The supply and demand constraints of cargo volume are as follows: S d ≥P d ,in, That is, the actual supply of each cargo transportation demand must not be less than the planned supply.

[0068] Delivery deadline constraints, specifically: That is, the cargo delivery time at each station must meet the delivery deadline constraints and must not exceed the latest delivery time.

[0069] Route skylight time constraints, specifically: That is, considering that the heavy-load railway skylight is a daily maintenance skylight, there must be no train running during this period of time every day, that is, it is necessary to ensure that the departure and arrival times of each train are outside the skylight period.

[0070] Locomotive quantity constraints, specifically: That is, each train in the group is pulled by a fixed number of locomotives, and the number of traction locomotives used on the entire line in a single day must not exceed the number of reserve locomotives on the entire line.

[0071] Step S3: solving the group train operation strategy corresponding to the Pareto optimality in the target optimization model based on a simulated annealing algorithm; wherein the number of trains in each group is the same and greater than or equal to 2.

[0072] Preferably, this step specifically includes:

[0073] (31) Initial solution generation, the generation rules are as follows:

[0074] Step 311. Randomly generate (M, N) within a specific range 2 The binary combination of the values ​​is decoded to obtain the decimal value (M, N), that is, the initial network operation group M and the number of trains N in the group are obtained.

[0075] Step 312. Generate the group's entire arrival 0-1 variable two-dimensional matrix v based on the initial (M, N), and randomly generate the group train's scattered arrival 0-1 two-dimensional matrix τ based on v.

[0076] In this step, a two-dimensional matrix v (v has M rows and I-1 columns; if v(m,i) = 1, it indicates that the entire m-th group arrives at station i, and each row of v contains at most one 1) is randomly generated based on v. A 0-1 two-dimensional matrix τ (τ is a 0-1 matrix with M rows and I-1 columns; if τ(m,i) = 1, it indicates that the m-th group has a train arriving at station i, and the number of rows in τ containing 1 elements is greater than 1 and less than or equal to N) representing the scattered arrival of trains in the group is generated. The two-dimensional matrices v and τ represent the specific information of the entire group and the scattered arrival of trains in the group at a certain station, respectively. The main relationship between the two matrices is: for group m and terminal station i, if the matrix v(m,i) = 1, then τ(m,:) = 0 (i.e., all elements in the m-th row of the τ matrix are 0); if the matrix v(m,i) = 0, then the m-th row of τ contains more than 1 and less than or equal to N 1 elements.

[0077] Step 313. Based on the two-dimensional matrices v and τ representing the group's entire or scattered arrival information in Step 312, generate a three-dimensional matrix γ representing the specific arrival information of each train in the group, considering that the departure time of the first train in the group is the window end time TC e , combined with the group departure time and the train departure time t1 and t2 within the group, generate the group train departure time two-dimensional matrix T start , based on the relevant parameters {l a ,L d ,V normal ,V away ,d,U m,n} and decision variables Calculate the travel time T of various groups of arrival type trains and dispersed arrival type group trains respectively m and T' m,n ,in, Generate group train arrival schedule T end .

[0078] In this step, the three-dimensional matrix γ is represented by I-1 as the number of three-dimensional matrix pages, M as the number of two-dimensional matrix rows of each page, and N as the number of two-dimensional matrix columns of each page. γ(m,n,i)=1 means that the terminal station of train n in group m is i. end The elements in the matrix are represented as Refers to the arrival time of the nth train of the mth group.

[0079] Step 314. The above operation strategy (γ, Tstart ,T end ) complies with the supply and demand constraints of goods quantity S d ≥P d , delivery deadline constraints Route skylight time constraints and Locomotive quantity constraints That is, it is used as the initial solution, otherwise return to Step 311 and regenerate.

[0080] (32) Determine the energy function minZ = g·E+j·R, where E and R represent the economic cost and transportation time, respectively, g and j represent the weights corresponding to the economic cost and transportation time, respectively, and minZ indicates that the optimization objective of the target optimization model is to minimize the weighted sum of the economic cost and the total in-transit transportation time of the goods.

[0081] (33) Generation of Neighborhood Solutions

[0082] (M,N) generated for the initial solution 2 The binary combination is probabilistically perturbed, and a new combination (M, N) is obtained by decoding, and it is judged whether the new combination meets the constraint conditions; if so, then randomly generate according to Step 312-Step 314 of (31) and output the neighborhood solution; if not, the original (M, N) is subjected to random perturbation. 2 The binary combination is probabilistically perturbed again until the constraints are fully satisfied and the neighborhood solution is then sought. The iterative solution algorithm for the neighborhood solution is as follows:

[0083] Step 331. Determine the initial temperature T0, record the current cooling times k=0, and go to Step 332.

[0084] Step 332. Let the number of iterations be n=0. At the initial temperature, generate the initial strategy solution X0 according to the methods in (31)-(32) above and calculate the energy function value, then go to Step 333.

[0085] Step 333. After k cooling times, at temperature T k Next, let the solution after the n-1th iteration be X n-1 , in order to obtain the neighborhood solution X n , randomly update the MN numerical combination, calculate the corresponding function value energy Z(Xx), and go to Step 334.

[0086] Step 334. Use the Metropolis criterion to solve the current strategy n Test, if Z(X n )-Z(X n-1 )<0, then accept X unconditionallyn Instead of X n-1 ; If Z(X n )-Z(X n-1 )>0, then with probability γ n (T k ) accepts the neighborhood solution; γ n (T k )The solution formula is:

[0087]

[0088] Go to Step 335.

[0089] Step 335. At the corresponding temperature, set the upper limit of the number of iteration steps to When the number of iterations at the temperature reaches the upper limit, go to Step 336; otherwise, execute Step 333.

[0090] Step 336. Algorithm convergence termination judgment: The set termination criterion is that the external temperature is lower than the specified threshold. If the criterion is met, the algorithm is terminated. Otherwise, go to Step 337.

[0091] Step 337. Cool down according to the following formula:

[0092] T k+1 =p·Tk

[0093] After cooling at the Tx temperature, the temperature is updated. At this time, the number of cooling times is recorded as k=k+1, the number of iteration steps is set to n=0, and the process returns to Step 333. The loop iteration is performed at the new temperature, and finally the optimal operation strategy is obtained under the set temperature and the internal iteration number.

[0094] Preferably, in the cooling formula of Step 337, p takes a value of 0.98.

[0095] Furthermore, this embodiment uses the Shuohuang Railway as a reference for case study. The Shuohuang line has 33 stations. However, for the convenience of analysis, it is simplified into 7 support stations, which are numbered Y1-Y7. The line situation is as follows: Figure 4 shown.

[0096] After solving the above algorithm steps, the comparison of model target values ​​is shown in Table 1 below:

[0097] Table 1: Comparison of target values

[0098]

[0099] Table 1 shows that the multi-objective function values ​​under the conventional operation strategy are much higher than those under the group operation strategy. The conventional operation strategy, with a load of 5,000 tons per run, provides a better match between supply and demand and results in lower variable loss costs, but its fixed operation costs are higher than those under the group operation strategy. The total in-transit time for cargo under the group operation strategy is 196.945 hours, while that under the conventional operation strategy is 333.28 hours, a saving of 136.335 hours. This shows that while both operation strategies can achieve delivery within the delivery deadline, the group operation strategy better meets cargo timeliness requirements and can transport more cargo at a lower economic cost, improving the economic benefits for both the railway and the cargo owner.

[0100] Refer to above Figure 4 The following table 2 shows the relevant parameter settings of each station on the line, and table 3 shows the settings of other parameters of the line:

[0101] Table 2: Related parameters of the fulcrum station

[0102]

[0103] Table 3: Other line parameters

[0104]

[0105] The original data of each transportation demand is obtained through the support station parameters. The following table 4 shows the original data of transportation demand:

[0106] Table 4: Raw data on transportation demand

[0107] Transportation Demand Distance between starting and ending stations / km Delivery deadline Cargo demand / 10,000 tons Site Level d1 16 18:00 6 3 d2 137 19:00 7 2 d3 267 20:00 7 2 d4 434 21:00 8 1 d5 630 22:00 6 3 d6 811 23:00 8 2

[0108] Because the present invention uses a cost-type indicator data normalization formula, it needs to meet the requirements that the longer the distance, the more urgent the delivery time, the greater the demand for goods, and the higher the site level, the more important the transportation demand is. Therefore, the delivery deadline and customer level indicator data are adjusted in reverse order.

[0109] Table 5 below shows the adjusted transport demand indicator data:

[0110] Table 5 Transport demand index data (after adjustment)

[0111] Transportation Demand Distance between starting and ending stations / km Transport to level Cargo demand / 10,000 tons Customer Level d1 16 6 6 1 d2 137 5 7 2 d3 267 4 7 2 d4 434 3 8 3 d5 630 2 6 1 d6 811 1 8 2

[0112] The grey comprehensive evaluation method is used to obtain the comprehensive importance of transportation demand ω=[0.5387,0.5695,0.5229,0.8135,0.3996,0.6927].

[0113] The comprehensive importance of transport demand ω is introduced, combined with the solution algorithm of the present invention, and solved based on the above (33). The algorithm solves the optimal operation strategy for this example to be 11 groups, with 8 trains per group. The following table shows the optimal operation strategy for the group trains obtained by the solution (Q1-1 in the table below represents the first train of the first group):

[0114] Table 6: Group train operation strategy

[0115]

[0116]

[0117] Table 6 (continued)

[0118] Train number Operation path Train number Operation path Train number Operation path Train number Operation path Q7-5 Y1-Y2 Q7-6 Y1-Y2 Q7-7 Y1-Y2 Q7-8 Y1-Y2 Q8-1 Y1-Y7 Q8-2 Y1-Y7 Q8-3 Y1-Y5 Q8-4 Y1-Y5 Q8-5 Y1-Y4 Q8-6 Y1-Y3 Q8-7 Y1-Y3 Q8-8 Y1-Y2 Q9-1 Y1-Y4 Q9-2 Y1-Y4 Q9-3 Y1-Y4 Q9-4 Y1-Y4 Q9-5 Y1-Y4 Q9-6 Y1-Y4 Q9-7 Y1-Y4 Q9-8 Y1-Y4 Q10-1 Y1-Y7 Q10-2 Y1-Y7 Q10-3 Y1-Y7 Q10-4 Y1-Y7 Q10-5 Y1-Y7 Q10-6 Y1-Y7 Q10-7 Y1-Y7 Q10-8 Y1-Y7 Q11-1 Y1-Y2 Q11-2 Y1-Y2 Q11-3 Y1-Y3 Q11-4 Y1-Y3 Q11-5 Y1-Y4 Q11-6 Y1-Y5 Q11-7 Y1-Y6 Q11-8 Y1-Y7

[0119] Table 7: Group train departure times

[0120]

[0121]

[0122] Table 7 (continued)

[0123] Train number Departure time Arrival time Train number Departure time Arrival time Q9-3 10:12 12:53 Q9-4 10:18 12:59 Q9-5 10:24 13:05 Q9-6 10:30 13:11 Q9-7 10:36 13:17 Q9-8 10:42 13:23 Q10-1 11:00 19:07 Q10-2 11:06 19:13 Q10-3 11:12 19:19 Q10-4 11:18 19:25 Q10-5 11:24 19:31 Q10-6 11:30 19:37 Q10-7 11:36 19:43 Q10-8 11:42 19:49 Q11-1 12:00 20:07 Q11-2 12:06 18:54 Q11-3 12:12 16:58 Q11-4 12:18 15:18 Q11-5 12:24 14:05 Q11-6 12:30 14:11 Q11-7 12:36 12:48 Q11-8 12:42 12:54

[0124] Table 8 below shows the comparison of cargo supply under the two operation strategies:

[0125] Table 8: Comparison of cargo supply under the two operation strategies

[0126]

[0127] In this example, the model objective value under the conventional operation strategy is much higher than that under the group operation strategy, resulting in higher economic costs. Although both operation strategies can arrive within the latest delivery deadline, the group operation strategy better meets cargo timeliness requirements and can transport more cargo at a lower economic cost, thereby improving the economic benefits for both the railway and the cargo owner.

[0128] Example 2

[0129] Corresponding to the above embodiment, this embodiment discloses a heavy-load train operation strategy optimization system based on group control technology, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the computer program.

[0130] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A heavy-load train operation strategy optimization method based on group control technology, characterized in that: include: Step S1: Determine a dedicated heavy-haul railway line to be optimized that has the characteristic of "directly departing from a technical station"; Step S2: Based on the supply-demand relationship, delivery time constraints, window time, number of locomotives, and the comprehensive importance of transportation demand, a target optimization model is constructed to minimize the weighted sum of economic cost and total in-transit transportation time of the goods; Step S3: solving the group train operation strategy corresponding to the Pareto optimality in the target optimization model based on a simulated annealing algorithm; wherein the number of trains in each group is the same and greater than or equal to 2; The step S2 comprises: Step S21: Set the following parameters related to the target optimization model: Gather at the network stations, is the station number, where ; is the set of road network segments, is the segment number, where ; is the length of segment a; is the number of groups opened on a single day, is the group number, where ; is the number of unit heavy-load trains in the group, is the train number, where ; is the transport demand set, is the transport requirement number, where ; Expressing transportation demand The actual supply, Expressing transportation demand Theoretical demand; Indicates the starting and ending points are and section; Indicates the normal operating speed of the train; Indicates the train disassembly and running speed; It represents the unit operation cost; represents the fixed cost of running a train at the first station; Indicates the departure interval between groups; Indicates the departure interval of each train in the group; Indicates the skylight start time. Indicates the end time of the skylight; Indicates the number of traction locomotives required for a unit train; Indicates the maximum number of traction locomotives in the road network; represents the departure time of train n in group m, represents the arrival time of train n in group m; Indicates the load of a unit heavy-load train; Indicates the comprehensive importance of transportation demand d; The transport distance representing the transport demand d; represents the transit time of the entire group m, represents the en route running time of train n of group m with dispersed arrivals; The relevant decision variables include: If the station For transportation needs The terminal is 1, otherwise it is 0; It indicates that if segment a belongs to the path ending at station i, it is 1, otherwise it is 0; If the group Train Transportation needs If yes, it is 1, otherwise it is 0; Indicates that if the group The entire group transportation demand is If yes, it is 1, otherwise it is 0; Step S22: The model optimization objective is divided into two parts: economic cost and cargo transportation time. The economic cost is divided into the departure cost at the originating technical station and the loss cost under supply and demand matching. The departure cost at the originating technical station is expressed as ; The loss cost under supply and demand matching is expressed as ,in, The total transit time of freight is composed of the total transit time of the entire group of arrival type trains, the total departure interval time, and the total transit time of the dispersed arrival type trains. The specific function is expressed as: ; Step S23: Set the following constraints on the target optimization model: Constraints on the supply and demand of cargo volume, specifically: ,in, ; Delivery deadline constraints are as follows: ; Route skylight time constraints, specifically: , ; Locomotive quantity constraints, specifically: .

2. The method according to claim 1, characterized in that The step S3 comprises: (31) The initial solution is generated according to the following rules: Step 311. Randomly generate within a specific range Binary combination of values ​​and decoding to obtain decimal value , that is, obtain the initial road network opening group and the number of trains in the group ; Step 312. Based on the initial ( Generate a two-dimensional matrix of 0-1 variables for the entire group ,based on Randomly generate group trains to disperse arrival 0-1 two-dimensional matrix ; Step 313. Based on the two-dimensional matrix representing the group or scattered arrival information in Step 312 and , generating a three-dimensional matrix that can represent the specific arrival information of each train in the group , considering the departure time of the first group is the end time of the skylight , combined with the group departure time and the departure time of the trains within the group and , generate a two-dimensional matrix of group train departure times , based on relevant parameters and decision variables , calculate the travel time of various groups of arrival type trains and dispersed arrival type group trains respectively and ,in, , , generate group train arrival schedule ; Step 314. The above opening strategy ( ) Comply with the supply and demand constraints of goods volume , delivery deadline constraints , Line skylight time constraints and , locomotive quantity constraints , that is, used as the initial solution, otherwise return to Step 311 to regenerate; (32) Determine the energy function ,in, and represent economic cost and transportation time respectively, and Indicates the weights corresponding to economic cost and transportation time respectively, It indicates that the optimization objective of the target optimization model is to minimize the weighted sum of economic cost and total in-transit transportation time of goods; (33) Generation of neighborhood solutions The initial solution generated Binary combination is probabilistically perturbed and decoded to obtain New combination, and judge whether the new combination meets the constraint conditions; if it meets the constraint conditions, then randomly generate according to Step 312-Step 314 of (31) and output the neighborhood solution; if it does not meet the constraint conditions, the original The binary combination is probabilistically perturbed again until the constraints are fully satisfied and the neighborhood solution is then sought. The iterative solution algorithm for the neighborhood solution is as follows: Step 331. Determine the initial temperature , record the current cooling times k = 0, and go to Step 332; Step 332. Let the number of iterations be n = 0. At the initial temperature, generate the initial strategy solution according to the method in (31)-(32) above. Calculate the energy function value and go to Step 333; Step 333. After k cooling times, at the temperature Next, the solution after the n-1th iteration is , in order to obtain the neighborhood solution , randomly update the MN numerical combination and calculate the corresponding function value energy , go to Step 334; Step 334. Use the Metropolis criterion to solve the current strategy Conduct inspection, if , then unconditionally accept replace ; like , then with probability Accept the neighborhood solution; The solution formula is: Go to Step 335; Step 335. At the corresponding temperature, set the upper limit of the number of iteration steps to When the number of iterations at the temperature reaches the upper limit, go to Step 336; otherwise, go to Step 333; Step 336. Algorithm convergence termination judgment: The set termination criterion is that the external temperature is lower than the specified threshold. If the criterion is met, the algorithm ends. Otherwise, go to Step 337; Step 337. Cool down according to the following formula: After the temperature is cooled, the temperature is updated. At this time, the number of cooling times k = k + 1 is recorded, the number of iteration steps n = 0 is set, and the process returns to Step 333. The loop iteration is performed at the new temperature, and the optimal operation strategy is finally obtained under the set temperature and the internal iteration number.

3. The method according to claim 2, characterized in that In the cooling formula of Step 337, The value is 0.

98.

4. The method according to any one of claims 1 to 3, characterized in that: The comprehensive importance of the transportation demand is determined based on the delivery deadline, the distance between two stations, the demand for goods and the customer level.

5. The method according to claim 4, characterized in that The comprehensive importance of the transportation demand is obtained by the grey comprehensive evaluation method.

6. A heavy-load train operation strategy optimization system based on group control technology, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

Images