CN115081119B - Method, device and equipment for optimizing train loading and readable storage medium - Google Patents

Method, device and equipment for optimizing train loading and readable storage medium Download PDF

Info

Publication number
CN115081119B
CN115081119B CN202210851160.1A CN202210851160A CN115081119B CN 115081119 B CN115081119 B CN 115081119B CN 202210851160 A CN202210851160 A CN 202210851160A CN 115081119 B CN115081119 B CN 115081119B
Authority
CN
China
Prior art keywords
information
algorithm
cargo
loading
result
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202210851160.1A
Other languages
Chinese (zh)
Other versions
CN115081119A (en
Inventor
刘�文
王海潮
孙逊
兰建华
高慕瑾
陈彦如
孙雪松
孙西敬
张苏波
刘斌
王充
张梦琪
王博
张法铭
何倩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southwest Jiaotong University
China Railway Siyuan Survey and Design Group Co Ltd
Original Assignee
Southwest Jiaotong University
China Railway Siyuan Survey and Design Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southwest Jiaotong University, China Railway Siyuan Survey and Design Group Co Ltd filed Critical Southwest Jiaotong University
Priority to CN202210851160.1A priority Critical patent/CN115081119B/en
Publication of CN115081119A publication Critical patent/CN115081119A/en
Application granted granted Critical
Publication of CN115081119B publication Critical patent/CN115081119B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • Human Resources & Organizations (AREA)
  • General Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Computer Hardware Design (AREA)
  • Quality & Reliability (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Operations Research (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Automation & Control Theory (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)

Abstract

The invention provides an optimization method, a device, equipment and a readable storage medium for train loading, which relate to the technical field of logistics and comprise the steps of obtaining cargo information and compartment information, wherein the compartment information comprises the maximum bearing weight information and the maximum bearing volume information of a compartment; constructing a calculation model of express train loading considering cargo conflict relationship constraints; and taking the cargo information and the carriage information as input information of the calculation model to obtain the minimum value of the number of the carriages, and recording the minimum value as the optimization result of the express train loading. The invention has the beneficial effects that: an integer programming model related to the train loading problem is expanded; a multi-arm gambling machine algorithm in reinforcement learning is adopted, a heuristic algorithm framework based on the multi-arm gambling machine is designed, and a proper algorithm is automatically selected to solve a calculation model according to the historical performance of the heuristic algorithm, so that the solving efficiency can be obviously improved, and higher solving quality can be obtained with lower calculation cost.

Description

Method, device and equipment for optimizing train loading and readable storage medium
Technical Field
The invention relates to the technical field of logistics, in particular to a method, a device and equipment for optimizing train loading and a readable storage medium.
Background
With the revolution of the industrial structure in China and the growth of the e-commerce industry, the transportation requirements of small batch, multiple categories, high value and high timeliness are gradually increased, and railway freight begins to change to modern logistics. In order to improve the railway logistics distribution efficiency, a fast freight train is opened for railway freight, and fast freight transportation is gradually developed, wherein loading is the most basic operation link for fast freight train freight. The efficient and reasonable loading method can improve the utilization rate of the carriage, reduce the transportation cost, meet logistics demand response more finely and improve customer satisfaction.
In the freight practice of express trains, the complexity of cargo loading is increased due to the variety of categories of customer orders. In addition to the weight and volume of the cargo, the different properties of the transported cargo, such as hygiene, volatility, fragility and safety, result in some cargo not being allowed to be loaded into a compartment at the same time, such as food, medicine and chemicals, i.e. there is a conflict relationship between the cargo. Therefore, the invention aims to find a loading scheme of the express train, which minimizes the number of carriages, under the condition of comprehensively considering the weight, the volume and the conflict relation constraint of the goods.
Disclosure of Invention
The invention aims to provide a method, a device, equipment and a readable storage medium for optimizing train loading, so as to improve the problems. In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
in a cargo loading model of a rapid transit train, most of researches in the prior art only consider weight and volume constraints of cargos and do not take conflict relations of the cargos into consideration, and the conflict relation constraints of the cargos are considered, so that the cargo loading model is more in line with the actual logistics scene of the rapid transit train; 2. in the aspect of algorithm design, no method for solving the loading problem of the express train in consideration of conflict relationship constraint exists at present, but a single method for designing the loading problem of other trains is difficult to balance the solving quality and the solving efficiency, and particularly, for large-scale complex loading problems, a better loading scheme is difficult to provide in a shorter solving time. And a meta-heuristic algorithm-based meta-heuristic selection method, such as a tabu algorithm, adds a heuristic algorithm used in the last iteration to a tabu table; in each iteration of the evolutionary algorithm, a new heuristic algorithm is selected through operations such as crossing and mutation, and the new heuristic algorithm is not intelligently selected according to the performance of the algorithm in different stages of different calculation examples.
The invention relates to a method for solving the loading problem of a rapid transit train, which simultaneously considers the weight, the volume and the conflict relationship constraint of goods. In order to overcome the defects, an integer programming model for express train loading is constructed; a heuristic algorithm frame based on a multi-arm gambling machine is adopted for solving, so that a express train loading scheme with the minimum carriage number under the condition that constraint conditions are met is obtained.
In a first aspect, the present application provides a method for optimizing train loading, including:
acquiring cargo information, wherein the cargo information comprises weight information, volume information and conflict relation information of each cargo in a cargo set loaded on a train, and the cargo is to be loaded on a express train;
obtaining compartment information, wherein the compartment information comprises maximum bearing weight information and maximum bearing volume information of a compartment, and the compartment is a compartment to be loaded with the goods;
constructing a calculation model of the express train loading considering the constraint of the cargo conflict relationship;
and taking the cargo information and the carriage information as input information of the calculation model, solving the calculation model to obtain the minimum value of the number used by the carriages, and recording the minimum value as an optimization result of the express train loading.
In a second aspect, the present application further provides an optimization device for train loading, including a first obtaining module, a second obtaining module, a constructing module and a solving module, wherein:
a first acquisition module: the system comprises a cargo information acquisition module, a cargo information processing module and a cargo information processing module, wherein the cargo information acquisition module is used for acquiring cargo information, the cargo information comprises weight information, volume information and conflict relationship information of each cargo in a cargo set loaded on a train, and the cargo is to be loaded into a express train;
a second obtaining module: the system comprises a load bearing device, a load bearing device and a load bearing device, wherein the load bearing device is used for obtaining carriage information, the carriage information comprises maximum bearing weight information and maximum bearing volume information of a carriage, and the carriage is a carriage to be loaded with goods;
constructing a module: a calculation model for building the express train loading that takes into account cargo conflict relationship constraints;
a solving module: and the system is used for solving the calculation model by taking the cargo information and the carriage information as input information of the calculation model to obtain the minimum value of the number of the carriages, and recording the minimum value as the optimization result of the express train loading.
In a third aspect, the present application further provides an optimization device for train loading, including:
a memory for storing a computer program;
a processor for implementing the steps of the method for optimizing train loading when executing the computer program.
In a fourth aspect, the present application further provides a readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the optimization method based on train loading.
The invention has the beneficial effects that: the invention starts from a physical logistics scene, constructs a rapid transit train loading model considering weight, volume and conflict relation constraint of cargos simultaneously, and expands an integer planning model related to train loading problems. A heuristic algorithm frame based on the multi-arm gambling machine is designed, and the solving efficiency and the solving quality are effectively improved. Specifically, a multi-arm gambling machine algorithm framework is designed to solve the calculation model, the multi-arm gambling machine algorithm framework can adaptively select a proper algorithm to solve the calculation model according to the historical performance of heuristic algorithms, and the solution efficiency can be remarkably improved. Three heuristic algorithms are specially designed according to the characteristics of the calculation model, wherein the three heuristic algorithms comprise an improved self-adaptive large-scale neighborhood searching algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm. The multi-arm gambling machine algorithm framework expands the solving technology of the heuristic algorithm and can obtain higher solving quality with smaller calculation cost.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the embodiments of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic flow chart of an optimization method for train loading according to an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of an optimization device for train loading according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an optimization device for train loading according to an embodiment of the present invention.
In the figure: 701. a first acquisition module; 702. a second acquisition module; 703. building a module; 704. a solving module; 7041. a building unit; 7042. selecting a unit; 7043. a first calculation unit; 7044. a setting unit; 7045. an input unit; 70451. calculating a profit unit; 704511, and a profit judgment unit; 704512, an iteration unit; 70452. an update unit; 70453. a selection unit; 70454. a result solving unit; 7046. a processing unit; 7047. an extension unit; 7048. a first judgment unit; 7049. a second judgment unit; 7050. a third judgment unit; 7051. a fusion unit; 800. optimizing equipment for train loading; 801. a processor; 802. a memory; 803. a multimedia component; 804. an I/O interface; 805. a communication component.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
Example 1:
the problem of loading the express trains considering the constraint of the conflict relationship is a ubiquitous problem in a railway logistics transportation scene, and the constraint of the weight, the volume and the conflict relationship of cargos are simultaneously considered. This problem extends the train loading problem considering only cargo weight and volume constraints, which are more challenging due to the complex constraints. At present, almost no method for solving the loading problem of the express train considering the constraint of the goods conflict relationship exists. The method for solving other loading problems mainly comprises the following steps: heuristic algorithms including tabu search algorithm, local search algorithm, simulated annealing algorithm, etc.; and precise algorithms including branch-and-bound algorithms, branch pricing algorithms, heuristic algorithms, and the like.
However, in the methods, a single algorithm often cannot achieve balance in terms of solving quality and solving efficiency at the same time, and particularly for large-scale complex loading problems, it is more difficult to provide a better express train loading scheme within reasonable solving time. And a meta-heuristic algorithm-based meta-heuristic selection method, such as a tabu algorithm, adds a heuristic algorithm used in the last iteration to a tabu table; in each iteration of the evolutionary algorithm, a new heuristic algorithm is selected through operations such as crossing and mutation, and the new heuristic algorithm is not intelligently selected according to the performance of the algorithm in different stages of different calculation examples. The invention designs a heuristic algorithm framework based on the multi-arm gambling machine, can adaptively select a proper algorithm to solve a calculation model according to the historical performance of the heuristic algorithm, and can obviously improve the solving efficiency. Three heuristic algorithms are specially designed according to the characteristics of the calculation model, wherein the three heuristic algorithms comprise an improved self-adaptive large-scale neighborhood searching algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm. The multi-arm gambling machine algorithm framework expands the solving technology of the heuristic algorithm and can obtain higher solving quality with smaller calculation cost.
The embodiment provides an optimization method for train loading.
Referring to fig. 1, it is shown that the method includes step S100, step S200, step S300 and step S400.
S100, acquiring cargo information, wherein the cargo information comprises weight information, volume information and conflict relation information of each cargo in a cargo set loaded by a train, and the cargo is to be loaded into the express train;
it will be appreciated that in this step, a set is givennCollection of individual goods
Figure 238918DEST_PATH_IMAGE001
Each goods item
Figure 156059DEST_PATH_IMAGE002
All have three properties of weight
Figure 935796DEST_PATH_IMAGE003
Volume, volume
Figure 331005DEST_PATH_IMAGE004
And a cargo conflict relationship diagram
Figure 207694DEST_PATH_IMAGE005
Is represented by
Figure 30157DEST_PATH_IMAGE006
The conflict relationship between all the goods in which the vertex sets
Figure 297190DEST_PATH_IMAGE006
For collections of goods, edge collectionsECorresponding to the conflict relationship between the goods.
S200, obtaining compartment information, wherein the compartment information comprises maximum bearing weight information and maximum bearing volume information of a compartment, and the compartment is a compartment to be loaded with goods;
it will be appreciated that in this step, each car of the same type has two attributes: weight limitationWAnd volume limitationV。Any goods can be loaded into the same compartment as long as they do not conflict with each other and the sum of their weight and volume does not exceed the weight and volume limits of the compartment.
S300, constructing a calculation model of the express train loading considering the constraint of the goods conflict relationship.
It can be understood that, in this step, an integer planning model for the loading problem of the express train is constructed while considering the weight, volume and conflict relationship constraints of the cargo, which is specifically as follows:
Figure 496090DEST_PATH_IMAGE008
(1)
Figure 227286DEST_PATH_IMAGE010
(2)
Figure 220650DEST_PATH_IMAGE011
， (3)
Figure 974979DEST_PATH_IMAGE012
， (4)
Figure 711991DEST_PATH_IMAGE013
，(5)
Figure 297693DEST_PATH_IMAGE014
， (6)
Figure 727538DEST_PATH_IMAGE015
， (7)
wherein the content of the first and second substances,Irepresenting a collection of goods, each good
Figure 969163DEST_PATH_IMAGE016
There are three attributes of each of the two types of media,
Figure 244287DEST_PATH_IMAGE003
the weight of the goods is represented by,
Figure 887757DEST_PATH_IMAGE004
which is indicative of the volume of the cargo,Erepresenting conflicting relationships between the goods.KRepresenting a set of cars, each car
Figure 285241DEST_PATH_IMAGE017
Are all of the same type.WIndicating the weight limit of the vehicle compartment,Vrepresenting the volumetric limitation of the car.
Figure 14162DEST_PATH_IMAGE018
Is a variable of 0-1 and represents cargoiWhether or not to be installed in the carriagek。
Figure 92977DEST_PATH_IMAGE019
Is a variable of 0-1 and represents cargojWhether or not to be installed in the carriagek。
Figure 590954DEST_PATH_IMAGE020
Is a variable of 0-1, representing a carkWhether or not to be used.
S400, taking the cargo information and the carriage information as input information of the calculation model, solving the calculation model to obtain the minimum value of the number of the carriages, and recording the minimum value as the optimization result of the express train loading.
The objective function (1) in the above model represents minimizing the number of cars used. The constraint (2) ensures that each cargo can only be loaded into one car. Constraints (3) and (4) require that the total weight and volume of cargo loaded into the same car not exceed the weight and volume limits of the car. Constraint (5) indicates that two goods in conflict relationship cannot be loaded into the same car. Constraints (6) - (7) define decision variables.
Specifically, a calculation model of express train loading considering cargo conflict relation constraint is constructed, cargo information and carriage information are used as input information of the calculation model, and the calculation model is solved to obtain the minimum value of the carriage use number.
Preferably, S400 includes S401, S402, and S403:
s401, adopting a dobby machine algorithm in reinforcement learning, designing a framework based on a heuristic algorithm of the dobby machine, and solving the calculation model to obtain a solution result;
s402, based on the historical performance of the heuristic algorithms, preferentially selecting the most suitable algorithm in the heuristic algorithms as an optimal algorithm to carry out iterative solution, wherein the heuristic algorithms comprise an improved self-adaptive large-scale neighborhood search algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm;
and S403, calculating to obtain a first solving result according to the optimal algorithm and the solving result.
Specifically, cargo information and compartment information are input into a calculation model, and the calculation model is solved through a multi-arm gambling machine-based heuristic algorithm framework to obtain the minimum value of the number of used compartments. The algorithm flow is described as follows:
and the multi-arm gambling machine algorithm framework adaptively selects a proper algorithm to solve the calculation model according to the historical performance of the heuristic algorithm. Based on the classic dobby algorithm, the invention replaces random selection with the roulette algorithm and sets the upper and lower bounds of optimistic initial values and action values to increase the probability of algorithm exploration.
Preferably, S400 further comprises S404 and S405:
s404, setting an optimistic initial value, a probability value of greedy selection and a threshold value of action value;
s405, based on the optimistic initial value, the probability value selected by greedy and the threshold value of the action value, adopting
Figure 424918DEST_PATH_IMAGE021
And selecting a proper heuristic algorithm, using the goods information and the carriage information as input information of the calculation model, and solving the calculation model to obtain the first solving result.
In the dobby algorithm, each action has a desired benefit when selected, called the action value. In the invention, an action is to select an efficient heuristic algorithm to solve a calculation model, the benefit refers to the improvement degree of the heuristic algorithm on the current solution, and the action value is the expected benefit after the heuristic algorithm is selected.
Preferably, the step S405 further includes S4051, S4052, S4053 and S4054, where:
s4051, obtaining a target value according to the selected heuristic algorithm, and calculating the benefit of the heuristic algorithm according to the improvement degree of the current solution of the target value;
s4052, updating the action value according to the income;
s4053, adopt
Figure 375557DEST_PATH_IMAGE021
Selecting the heuristic algorithm with the maximum value by a greedy method, or selecting the heuristic algorithm for roulette with the updated action value by adopting a probability of 1-;
s4054, according to the selected heuristic algorithm, the cargo information and the carriage information are used as input information of the calculation model, and the calculation model is solved to obtain the first solving result.
Preferably, the step S4051 further includes steps S40511 and S40512, where:
s40511, judging whether the target value obtained by the selected heuristic algorithm improves the current solution, if so, increasing positive income in the action value, and updating the action value;
s40512, if the current solution is not improved, increasing negative income in the action value, updating the action value, and reselecting from the rest heuristic algorithms; and terminating the algorithm until all the heuristic algorithms cannot improve the current solution or reach a preset iteration number, and outputting a result.
At each iteration, adopt
Figure 258062DEST_PATH_IMAGE021
Greedy method selection value maximum heuristic, 1-
Figure 610546DEST_PATH_IMAGE021
The probability of (c) makes a roulette selection. And if the selected heuristic algorithm does not improve the current solution, reselecting from the rest heuristic algorithms until all heuristic algorithms are found and the current solution cannot be improved or the iteration times are reached.
The initialization parameters of the algorithm are as follows, the action is set to 3, i.e. three heuristic algorithms can be selected: an improved self-adaptive large-scale neighborhood search algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm. The optimistic initial value is set to 15 and the upper and lower bounds of the action value are set to 15 and 0.3, respectively. The action value update formula is as follows:
Figure 553094DEST_PATH_IMAGE022
， (8)
Figure 53346DEST_PATH_IMAGE023
， (9)
Figure 473963DEST_PATH_IMAGE024
， (10)
wherein the content of the first and second substances,
Figure 946532DEST_PATH_IMAGE025
heuristic algorithm for representative selectionjIn thattThe gain obtained at the time step is,
Figure 59982DEST_PATH_IMAGE026
representing the value of the objective function of the current solution,
Figure 50459DEST_PATH_IMAGE027
by heuristic algorithmsjIn thattThe objective function value obtained at the time step,
Figure 274767DEST_PATH_IMAGE028
representing heuristic algorithmsjIn thattMotion value estimated at time step,
Figure 336264DEST_PATH_IMAGE029
Representing heuristic algorithmsjIn thattThe action value of the +1 time step estimate.
Figure 620615DEST_PATH_IMAGE031
The step size is indicated. When the heuristic algorithm improves the current solution, updating the action value according to the formula (9); when the heuristic does not improve the current solution, the action value is updated according to equation (10). The action value cannot exceed the preset upper and lower bounds. And when the selected heuristic algorithm can not improve the current solution, selecting another heuristic algorithm for searching, and if the current solution can not be improved or the iteration times are reached, terminating the algorithm.
Specifically, the multi-arm gambling machine algorithm framework involves three heuristic methods as follows:
the improved self-adaptive large-scale neighborhood searching algorithm comprises the following steps: a plurality of groups of destruction operators and repair operators are designed, and the current solution is continuously improved. In each iteration, according to the historical performance, each damage operator and each repair operator are selected and weight adjusted, and the optimization capability of the algorithm is improved.
In the improved adaptive large-scale neighborhood search algorithm, the initial solution employs a simple loading strategy that easily satisfies constraints, i.e., one cargo is loaded into one car, respectively. The number of failure points was set to 0.3 cargo count. There are three destruction operators, which are:
(1) Minimal conflict cargo removal. The number of cars is randomly picked until the number of failure points is reached, and the most conflicting goods are removed in each car picked.
(2) A random cargo removal operator. The number of cars is randomly chosen until the number of failure points is reached, and then one cargo is randomly removed from each car.
(3) The maximum weight cargo is removed. The number of cars is randomly picked until the number of failure points is reached, and the heaviest weight of cargo is removed in each car picked.
(4) The maximum volume of cargo is removed. The number of cars is randomly picked until the number of failure points is reached, and the largest volume of cargo is removed in each car picked.
The repair operators are two, respectively:
(1) A weight greedy insertion operator. A previously deleted good is selected and inserted into a car if the remaining weight after insertion is minimal without violating the constraints.
(2) A volume greedy insertion operator. A previously deleted good is selected and inserted into a car if the remaining volume after insertion is minimal without violating the constraints.
If a heuristic algorithm is selected and a global optimal solution is obtained, updating the score of the operator according to the formula (11); if a heuristic algorithm is selected and the current optimal solution is obtained, updating the score of the operator according to the formula (12); if a heuristic is selected and results in a degraded solution, the score of the operator is updated according to equation (13);
Figure 95458DEST_PATH_IMAGE032
， (11)
Figure 123457DEST_PATH_IMAGE033
， (12)
Figure 305040DEST_PATH_IMAGE034
， (13)
wherein the content of the first and second substances,
Figure 494713DEST_PATH_IMAGE035
representing heuristic algorithmsjIn the first placet+1The fraction of the sub-iterations is,
Figure 394536DEST_PATH_IMAGE036
representing heuristic calculationsMethod of makingjIn the first placetFraction of the sub-iteration. If only better solution solutions than the current solution are accepted, it is easy to fall into local optima. Therefore, it is necessary to accept the differential solution with a certain probability, and the acceptance probability is set to beT，TThe updating is performed according to the following formula:
Figure 22963DEST_PATH_IMAGE037
， (14)
wherein the content of the first and second substances,Tindicating the probability of acceptance, the initial value is set to 0.2,iteras a result of the total number of iterations,
Figure 324631DEST_PATH_IMAGE038
is the number of times that has been currently iterated. The selection of the heuristic depends on its weight, on completiontAnd (4) performing iteration, updating the weight once, wherein the weight updating formula of the heuristic algorithm is as follows:
Figure 685206DEST_PATH_IMAGE039
， (15)
wherein, the first and the second end of the pipe are connected with each other,
Figure 72325DEST_PATH_IMAGE040
is shown astHeuristic algorithm during sub-iterationjThe weight of (a) is determined,
Figure 504443DEST_PATH_IMAGE041
denotes the firsttHeuristic Algorithm during +1 iterationjThe weight of (a) is determined,
Figure 395039DEST_PATH_IMAGE042
representing heuristic algorithmsjThe score of (a) is obtained,
Figure 192093DEST_PATH_IMAGE043
representing heuristic algorithmsjNumber of times selected in historical iterations.
Figure 800929DEST_PATH_IMAGE044
Is a weight parameterA number in the range of [0,1 ]]。
Improved max-min ant colony system algorithm: the improved maximum and minimum ant colony system algorithm adopted by the invention is used as one of the ant colony algorithms, and has the advantages that the upper and lower boundaries of pheromones are set, the searchability of the algorithm is increased, and the situation that the difference of the pheromones is too large to be locally optimal is prevented. The algorithm also adopts an elite strategy, and in the current iteration, the ants are allowed to release the pheromone only when a global optimal solution or a local optimal solution is obtained.
Concentration of pheromones in the algorithm
Figure 36738DEST_PATH_IMAGE045
Simulate ants to select goodsiAnd goodsjThe update formula of the pheromone residual of (2) is as follows:
Figure 47420DEST_PATH_IMAGE046
， (16)
wherein, the first and the second end of the pipe are connected with each other,
Figure 15376DEST_PATH_IMAGE047
is the volatilization coefficient of the pheromone, is set to 0.1, and the upper and lower bounds of the pheromone are set to 0.001 and 0.999 respectively if the goods are combinediAndjif it occurs in the historical optimal solution or the current optimal solution, then
Figure 111508DEST_PATH_IMAGE048
Is set to 1 and otherwise is 0.
Figure 823112DEST_PATH_IMAGE049
Is an articleiAndjpheromone concentration in between. The ants build the express train loading mode by continuously selecting the next goods, the selection mode comprises greedy selection and roulette selection, the probability of the greedy selection is set to be 0.3, namely, the greedy selection is carried out with the probability of 0.3, and the roulette selection is carried out with the probability of 0.7. The greedy selection is as follows:
Figure 750616DEST_PATH_IMAGE050
(17)
Figure 889474DEST_PATH_IMAGE051
， (18)
wherein, the first and the second end of the pipe are connected with each other,
Figure 472902DEST_PATH_IMAGE052
is an articleiAndqthe concentration of the pheromone in between,
Figure 722618DEST_PATH_IMAGE053
the weight and volume of the current cargo relative to the weight and volume of the next cargo are shown as heuristic factors, and the smaller the difference, the easier the selection.
Figure 770208DEST_PATH_IMAGE054
、
Figure 79967DEST_PATH_IMAGE055
The weight of the article is the weight of the article,
Figure 150691DEST_PATH_IMAGE056
、
Figure 204098DEST_PATH_IMAGE057
is the volume of the article.
Figure 106194DEST_PATH_IMAGE058
To satisfy the current constraint of an unloaded set of compatible shipments,
Figure 586854DEST_PATH_IMAGE059
、
Figure 879295DEST_PATH_IMAGE044
is a weight parameter.
Roulette was selected as follows:
Figure 736393DEST_PATH_IMAGE060
， (19)
wherein the content of the first and second substances,
Figure 430680DEST_PATH_IMAGE061
representing antsuIndicating selection of the next compatible itemjThe probability of (c).
Figure 135769DEST_PATH_IMAGE062
Is an articleiAndqthe concentration of the pheromone in between,
Figure 181085DEST_PATH_IMAGE049
is an articleiAndjpheromone concentration in between.
Figure 576294DEST_PATH_IMAGE063
、
Figure 125087DEST_PATH_IMAGE053
The weight and volume of the current cargo relative to the weight and volume of the next cargo are shown as heuristic factors, and the smaller the difference, the easier the selection.
Figure 275446DEST_PATH_IMAGE064
To satisfy the current constraint of an unloaded set of compatible shipments,
Figure 542479DEST_PATH_IMAGE059
、
Figure 741379DEST_PATH_IMAGE044
is a weight parameter.
Improved best-fit heuristic algorithm: and (3) sequentially loading the cargoes into the carriages, if the goods can meet the constraints of weight, volume and conflict relationship, loading the cargoes into the carriage with the highest score, and scoring in a rule formula (20), wherein after the current cargoes are added into the carriage, the more balanced the residual load and the residual volume, the higher the probability of being selected for loading. When goods cannot be loaded into the carriage which meets the constraint, a new carriage is started.
Figure 144679DEST_PATH_IMAGE065
(20)
Wherein the content of the first and second substances,
Figure 403622DEST_PATH_IMAGE066
is a carriagekThe higher the score, the greater the probability of selection.
Figure 220268DEST_PATH_IMAGE067
The remaining load added to the car for the current load,
Figure 957280DEST_PATH_IMAGE068
as goodsjThe remaining volume of the car is added to the car,
Figure 480665DEST_PATH_IMAGE069
the maximum load-bearing weight of the carriage,
Figure 910509DEST_PATH_IMAGE070
the maximum carrying volume of the carriage is provided,
Figure 948873DEST_PATH_IMAGE059
is a weight parameter.
Preferably, the constructing a computation model for loading the express train considering the constraint of the conflict relationship, taking the cargo information and the car information as input information of the computation model, and solving the computation model, includes: and constructing a calculation model for loading the express train in consideration of the constraint of the cargo conflict relationship, taking the cargo information and the carriage information as input information of the calculation model, and solving the calculation model to obtain the minimum value of the carriage use number.
Preferably, the constructing a calculation model for express train loading considering conflict relationship constraints, taking the cargo information and the car information as input information of the calculation model, and solving the calculation model to obtain the minimum value of the number of used cars includes: solving the calculation model through a multi-arm gambling machine-based heuristic algorithm framework to obtain the minimum value of the using number of the carriages.
Preferably, the constructing calculation model for express train loading considering conflict relationship constraint, taking the cargo information and the car information as input information of the calculation model, and solving the calculation model to obtain the minimum value of the number of used cars includes: the multi-arm gambling machine algorithm framework solves the calculation model, and adaptively selects the most appropriate one of three heuristic algorithms to solve according to the historical performance of the heuristic algorithms, wherein the most appropriate one of the three heuristic algorithms comprises an improved adaptive large-scale neighborhood search algorithm, an improved maximum-minimum ant colony system algorithm and an improved optimal adaptation algorithm, and the minimum value of the using number of the carriages is obtained.
Can be simply understood as: the dobby gambling machine algorithm framework comprises an improved adaptive large-scale neighborhood search algorithm, an improved maximum-minimum ant colony system algorithm and an improved best-fit algorithm. When solving the loading problem of the express train, the prior method does not adopt a multi-arm gambling machine algorithm framework to select a proper heuristic algorithm for solving, but the multi-arm gambling machine algorithm framework adopted by the invention can ensure that the algorithm obtains higher solving quality with lower calculation cost.
Specifically, solving the calculation model to obtain the minimum value of the number of used cars includes: through the dobby gambling machine algorithm framework, adopt
Figure 489575DEST_PATH_IMAGE021
The greedy method selects one of the three heuristic algorithms of the improved self-adaptive large-scale neighborhood search algorithm, the improved maximum and minimum ant colony system algorithm and the improved optimal adaptation algorithm with the highest action value to solve, and uses 1-
Figure 133046DEST_PATH_IMAGE021
The probability of (c) makes a roulette selection. When the selected heuristic has not improved the target value, the rest of the heuristics are calculatedAnd reselecting in the method until the target value cannot be improved or the preset iteration number is reached, and obtaining the minimum value of the using number of the carriages.
Preferably, S405 is preceded by S4051, S4052, S4053, S4054, S4055 and S4056, wherein:
s4051, preprocessing the collision cargo matrix loaded by the express train and the loading mode loaded by the express train according to a collision preprocessing algorithm to obtain a preprocessing result, wherein the preprocessing result comprises an expansion result and a loading result, and the preprocessing result comprises:
s4052, when the conflict cargo matrix loaded by the express train is preprocessed, extending an edge set in a conflict graph in the conflict cargo matrix to obtain an extension result;
s4053, if the goods are incompatible with other goods in the set, the goods can be loaded into a carriage independently to form a loading mode;
s4054, if the goods are compatible with only one goods in the set, loading the two goods into a carriage to form a loading mode;
s4055, if the goods are only compatible with two goods in the set and meet preset constraint conditions, loading the three goods into a carriage to form a loading mode; if the constraint is not satisfied, no processing is performed, and a loading result is obtained;
s4056, splicing and fusing the first solving result and the preprocessing result to obtain an optimized result of the express train loading.
In addition, it should be noted that, in the present invention, a conflict preprocessing algorithm is designed in the above steps, which specifically includes the following steps:
the conflict preprocessing algorithm comprises preprocessing of a conflict cargo matrix and preprocessing of a loading mode. The preprocessing of the collision cargo matrix helps to reduce the search space of the algorithm, and the preprocessing of the loading mode helps to reduce the number of cargoes needing to be processed.
When preprocessing the conflict cargo matrix, the conflict graph needs to be preprocessed
Figure 468213DEST_PATH_IMAGE071
Edge set in (1)EAnd (5) performing expansion. Firstly for any goods
Figure 259451DEST_PATH_IMAGE072
In aiAdding an edge between each corresponding goods with conflict relationship; second, for any two goods
Figure 338266DEST_PATH_IMAGE073
，
Figure 836243DEST_PATH_IMAGE074
There are cases whereiAndjan edge is added in between.
(1) GoodsiAndjincompatibility in terms of weight:
Figure 342311DEST_PATH_IMAGE075
(21)
(2) Goods (I)iAndjvolume incompatibility:
Figure 620845DEST_PATH_IMAGE076
(22)
in the formula (I), the compound is shown in the specification,Ia collection of goods is represented that is,
Figure 503351DEST_PATH_IMAGE077
the weight of the cargo is represented and,
Figure 855835DEST_PATH_IMAGE078
which is indicative of the volume of the cargo,Windicating the weight limit of the vehicle compartment,Vindicating the volumetric limitation of the car.
Pretreatment in a loading mode: after the preprocessing matrix is obtained, the following preprocessing can be performed on the initial express train loading mode:
(1) If the goods areiAnd in the collectionAre incompatible with other goods, the goods are not compatible with the other goodsiCan be independently loaded into a carriage to form a loading mode;
(2) If the goods areiOnly compatible with one cargo in the set, the two cargoes are loaded into a carriage to form a loading mode;
(3) If the goods areiAnd only two cargos in the collection are compatible, and the constraint condition is met, the three cargos are loaded into one carriage to form a loading mode. If the constraint is not satisfied, no processing is done.
The goods after the pretreatment do not enter the subsequent algorithm flow, but after the algorithm is finished, the loading modes and the solution found by the algorithm are spliced to obtain the final solution result. The preprocessing of the express train loading mode can effectively reduce the data scale needing to be processed.
Example 2:
as shown in fig. 2, the embodiment provides an express train loading device considering a cargo conflict relationship, which includes a first obtaining module 701, a second obtaining module 702, a constructing module 703 and a solving module 704, where:
the first obtaining module 701: the system comprises a cargo information acquisition module, a cargo information processing module and a cargo information processing module, wherein the cargo information acquisition module is used for acquiring cargo information, the cargo information comprises weight information, volume information and conflict relationship information of each cargo in a cargo set loaded on a train, and the cargo is to be loaded into a express train;
the second obtaining module 702: the system comprises a load bearing device, a load bearing device and a load bearing device, wherein the load bearing device is used for obtaining carriage information, the carriage information comprises maximum bearing weight information and maximum bearing volume information of a carriage, and the carriage is a carriage to be loaded with goods;
a building module 703: a calculation model for building the express train loading that takes into account cargo conflict relationship constraints;
the solving module 704: and the system is used for solving the calculation model by taking the cargo information and the carriage information as input information of the calculation model to obtain the minimum value of the number of the carriages, and recording the minimum value as the optimization result of the express train loading.
Preferably, the solving module 704 includes a constructing unit 7041, a selecting unit 7042 and a first calculating unit 7043, where:
construction unit 7041: the method comprises the steps that a multi-arm gambling machine algorithm in reinforcement learning is adopted, a framework based on a heuristic algorithm of the multi-arm gambling machine is designed, and a calculation model is solved to obtain a solution result;
selecting unit 7042: the optimal algorithm is selected from the heuristic algorithms preferentially to be used as an optimal algorithm for iterative solution based on the historical performance of the heuristic algorithms, and the heuristic algorithms comprise an improved self-adaptive large-scale neighborhood search algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm;
first calculation unit 7043: and calculating to obtain a first solving result according to the optimal algorithm and the solving result.
Preferably, the solving module 704 further includes a setting unit 7044 and an input unit 7045, where:
setting unit 7044: setting thresholds for optimistic initial values, probability values of greedy choices and action values;
input unit 7045: and the heuristic algorithm is selected based on the optimistic initial value, the probability value selected by greedy and the threshold value of the action value by adopting a greedy method and a roulette method, the goods information and the carriage information are used as input information of the calculation model, and the calculation model is solved to obtain the first solving result.
Preferably, first calculating unit 7043 further includes processing unit 7046, extending unit 7047, first determining unit 7048, second determining unit 7049, third determining unit 7050, and fusing unit 7051, where:
processing unit 7046: the system comprises a collision preprocessing algorithm, a collision cargo matrix loaded by the express train and a loading mode loaded by the express train are preprocessed according to the collision preprocessing algorithm to obtain a preprocessing result, wherein the preprocessing result comprises an expansion result and a loading result, and the collision cargo matrix comprises:
extension unit 7047: the system comprises a conflict graph database, a conflict graph database and a conflict graph database, wherein the conflict graph database is used for storing conflict graphs of the express trains;
first determining unit 7048: if the goods are incompatible with other goods in the collection, the goods can be loaded into a carriage independently to form a loading mode;
second determining unit 7049: for loading two said goods into a compartment, if said goods are compatible with only one said goods in the collection, to form a loading mode;
third determining unit 7050: if the goods are only compatible with two goods in the set and meet preset constraint conditions, loading the three goods into a carriage to form a loading mode; if the constraint is not satisfied, no processing is performed, and a loading result is obtained;
fusion unit 7051: and the first solving result and the preprocessing result are spliced and fused to obtain an optimization result of the express train loading.
Preferably, the input unit 7045 further includes a profit calculating unit 70451, an updating unit 70452, a selecting unit 70453, and a solution result unit 70454, where:
calculate revenue unit 70451: the target value is obtained according to the selected heuristic algorithm, and the income of the heuristic algorithm is calculated through the improvement degree of the current solution of the target value;
update unit 70452: updating the action value according to the income;
selection unit 70453: for selecting the heuristic algorithm with the highest value by using the greedy method, or using 1-
Figure 798383DEST_PATH_IMAGE021
Selecting the heuristic algorithm for roulette of the updated action value;
solution result unit 70454: and the calculation model is solved by taking the cargo information and the carriage information as input information of the calculation model according to the selected heuristic algorithm to obtain the first solving result.
Preferably, calculating profit cell 70451 includes determining profit cell 704511 and iterating cell 704512, where:
judge profit cell 704511: the heuristic algorithm is used for judging whether the target value obtained by the selected heuristic algorithm improves the current solution or not, if the current solution is improved, a positive gain is added to the action value, and the action value is updated;
iteration unit 704512: if the current solution is not improved, increasing negative income in the action value, updating the action value, and reselecting from the rest heuristic algorithms; and stopping the algorithm until all the heuristic algorithms cannot improve the current solution or reach the preset iteration times, and outputting a result.
The invention constructs a rapid transit train loading model considering weight, volume and conflict relation constraint of cargos simultaneously, and expands an integer planning model related to train loading problems. A heuristic algorithm framework based on the multi-arm gambling machine is designed to solve the calculation model, the multi-arm gambling machine framework can adaptively select a proper algorithm to solve the calculation model according to the historical performance of the heuristic algorithm, and the solving efficiency can be obviously improved. Three heuristic algorithms are specially designed according to the characteristics of the calculation model, wherein the three heuristic algorithms comprise an improved self-adaptive large-scale neighborhood searching algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm. The multi-arm gambling machine algorithm framework expands the solving technology of the heuristic algorithm, and can ensure that the algorithm obtains higher solving quality with lower computing cost.
It should be noted that, regarding the apparatus in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated herein.
Example 3:
corresponding to the above method embodiment, the present embodiment further provides a express train loading device considering a cargo conflict relationship, and the express train loading device considering a cargo conflict relationship described below and the express train loading method considering a cargo conflict relationship described above may be referred to in correspondence.
Fig. 3 is a block diagram of an express train loading device 800 in accordance with an exemplary embodiment showing a cargo conflict relationship. As shown in fig. 3, the express train loading apparatus 800 may include: a processor 801, a memory 802. The express train loading device 800 may also include one or more of a multimedia component 803, an i/O interface 804, and a communications component 805.
The processor 801 is configured to control the overall operation of the rapid train loading apparatus 800 to perform all or part of the steps of the rapid train loading method in consideration of the cargo conflict relationship. The memory 802 is used to store various types of data to support the operation of the express train loader 800, such data can include, for example, instructions for any application or method operating on the express train loader 800, as well as application-related data, such as contact data, transceived messages, pictures, audio, video, and the like. The Memory 802 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), electrically Erasable Programmable Read-Only Memory (EEPROM), erasable Programmable Read-Only Memory (EPROM), programmable Read-Only Memory (PROM), read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk or optical disk. The multimedia components 803 may include screen and audio components. Wherein the screen may be, for example, a touch screen and the audio component is used for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may further be stored in the memory 802 or transmitted through the communication component 805. The audio assembly also includes at least one speaker for outputting audio signals. The I/O interface 804 provides an interface between the processor 801 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication module 805 is used for wired or wireless communication between the express train loading unit 800 and other devices. Wireless communication, such as Wi-Fi, bluetooth, near Field Communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them, so that the corresponding communication component 805 may include: wi-Fi module, bluetooth module, NFC module.
In an exemplary embodiment, the express train loading Device 800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above-described express train loading method.
In another exemplary embodiment, a computer readable storage medium comprising program instructions that when executed by a processor implement the steps of the express train loading method described above is also provided. For example, the computer readable storage medium may be the memory 802 described above that includes program instructions that are executable by the processor 801 of the rapid train loading device 800 to perform the rapid train loading method described above.
Example 4:
corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment, and a readable storage medium described below and an express train loading method considering a cargo conflict relationship described above may be referred to in correspondence.
A readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the express train loading method of the above-described method embodiments.
The readable storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, an optical disk, or other various readable storage media capable of storing program codes.
In summary, the invention researches the express train loading problem considering cargo weight, volume and conflict relation constraint at the same time, firstly constructs an integer programming model of the express train loading problem considering cargo conflict relation constraint, and designs a heuristic algorithm framework based on a multi-arm gambling machine aiming at the model, wherein the algorithm can self-adaptively select the most appropriate one to solve according to the historical performance of the heuristic algorithm. In addition, a conflict preprocessing algorithm is designed for accelerating the algorithm, and the calculation time is reduced. The heuristic algorithm framework based on the multi-arm gambling machine can solve a loading scheme with higher quality for the loading problem of the express train in consideration of the weight, the volume and the conflict relationship constraint of goods in a shorter time.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of the changes or substitutions within the technical scope of the present invention, and shall cover the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (6)

1. A method for optimizing train loading, comprising:
acquiring cargo information, wherein the cargo information comprises weight information, volume information and conflict relation information of each cargo in a cargo set loaded on a train, and the cargo is to be loaded into a express train;
obtaining compartment information, wherein the compartment information comprises maximum bearing weight information and maximum bearing volume information of a compartment, and the compartment is a compartment to be loaded with the goods;
constructing a calculation model of the express train loading considering cargo conflict relationship constraints;
taking the cargo information and the carriage information as input information of the calculation model, solving the calculation model to obtain the minimum value of the number of the carriages, and recording the minimum value as an optimization result of the express train loading;
wherein, the solving the calculation model by using the cargo information and the carriage information as the input information of the calculation model comprises:
adopting a dobby gambling machine algorithm in reinforcement learning, designing a framework based on a heuristic algorithm of the dobby gambling machine, and solving the calculation model to obtain a solution result;
based on the historical performance of the heuristic algorithm, the most suitable algorithm in the heuristic algorithm is preferentially selected as the optimal algorithm to carry out iterative solution, wherein the heuristic algorithm comprises an improved self-adaptive large-scale neighborhood search algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm;
calculating to obtain a first solving result according to the optimal algorithm and the solving result;
wherein, the solving the calculation model by using the cargo information and the carriage information as the input information of the calculation model further comprises:
obtaining a target value according to the selected heuristic algorithm, and calculating the income of the heuristic algorithm according to the improvement degree of the current solution of the target value;
updating the action value according to the income;
selecting the heuristic algorithm with the maximum value by adopting an epsilon greedy method, or selecting the heuristic algorithm for the roulette with the updated action value by adopting a probability of 1-epsilon;
and solving the calculation model by taking the cargo information and the carriage information as input information of the calculation model according to the selected heuristic algorithm to obtain the first solving result.
2. The method for optimizing train loading according to claim 1, wherein the calculating a first solution result according to the optimization algorithm and the solution result further comprises:
preprocessing a collision cargo matrix loaded by the express train and a loading mode loaded by the express train according to a collision preprocessing algorithm to obtain a preprocessing result, wherein the preprocessing result comprises an expansion result and a loading result, and the preprocessing result comprises:
when the conflict cargo matrix loaded by the express train is preprocessed, expanding an edge set in a conflict graph in the conflict cargo matrix to obtain an expansion result;
if the goods are incompatible with other goods in the collection, the goods can be loaded into a carriage independently to form a loading mode;
if the goods are compatible with only one goods in the set, the two goods are loaded into a carriage to form a loading mode;
if the goods are only compatible with two goods in the set and meet preset constraint conditions, loading the three goods into a carriage to form a loading mode; if the constraint is not satisfied, no processing is performed, and a loading result is obtained;
and splicing and fusing the first solving result and the preprocessing result to obtain an optimized loading result of the express train.
3. An apparatus for optimizing train loading, comprising:
a first obtaining module: the system comprises a cargo information acquisition module, a cargo information processing module and a cargo information processing module, wherein the cargo information acquisition module is used for acquiring cargo information, the cargo information comprises weight information, volume information and conflict relationship information of each cargo in a cargo set loaded on a train, and the cargo is to be loaded into a express train;
a second obtaining module: the system comprises a load bearing device, a load bearing device and a load bearing device, wherein the load bearing device is used for obtaining carriage information, the carriage information comprises maximum bearing weight information and maximum bearing volume information of a carriage, and the carriage is a carriage to be loaded with goods;
constructing a module: a computational model for constructing the express train loading that takes into account cargo conflict relationship constraints;
a solving module: the system comprises a calculation model, a load optimization module and a load optimization module, wherein the load optimization module is used for calculating the load of the express train according to the load information and the carriage information;
wherein, the solving module comprises:
a construction unit: the method comprises the steps that a multi-arm gambling machine algorithm in reinforcement learning is adopted, a framework based on a heuristic algorithm of the multi-arm gambling machine is designed, and a calculation model is solved to obtain a solution result;
a selecting unit: the optimal algorithm is selected from the heuristic algorithms preferentially to be used as an optimal algorithm for iterative solution based on the historical performance of the heuristic algorithms, and the heuristic algorithms comprise an improved self-adaptive large-scale neighborhood search algorithm, an improved maximum and minimum ant colony system algorithm and an improved optimal adaptive algorithm;
the first calculation unit: the calculation module is used for calculating to obtain a first solving result according to the optimal algorithm and the solving result;
wherein, the solving module further comprises:
and a revenue calculation unit: the target value is obtained according to the selected heuristic algorithm, and the income of the heuristic algorithm is calculated through the improvement degree of the current solution of the target value;
an update unit: updating the action value according to the income;
a selection unit: selecting the heuristic algorithm with the highest value by using an epsilon greedy method, or selecting the heuristic algorithm for roulette of the updated action value by using a probability of 1-epsilon;
a solution result unit: and the calculation model is solved by taking the cargo information and the carriage information as input information of the calculation model according to the selected heuristic algorithm to obtain the first solving result.
4. The train loading optimization device of claim 3, wherein the first computing unit further comprises:
a processing unit: the system comprises a collision preprocessing algorithm, a collision cargo matrix and a loading mode, wherein the collision cargo matrix is loaded by the express train, the loading mode is loaded by the express train, and a preprocessing result is obtained, wherein the preprocessing result comprises an expansion result and a loading result, and the collision preprocessing algorithm comprises the following steps:
an extension unit: the system comprises a conflict graph database, a conflict graph database and a conflict graph database, wherein the conflict graph database is used for storing conflict graphs of the express trains;
a first judgment unit: the goods can be loaded into a carriage independently to form a loading mode if the goods are incompatible with other goods in the set;
a second judgment unit: for loading two said goods into a compartment, if said goods are compatible with only one said goods in the collection, to form a loading mode;
a third judging unit: if the goods are only compatible with two goods in the set and meet preset constraint conditions, loading the three goods into a carriage to form a loading mode; if the constraint is not satisfied, no processing is performed, and a loading result is obtained;
a fusion unit: and the first solving result and the preprocessing result are spliced and fused to obtain the optimized loading result of the express train.
5. An optimization apparatus for train loading, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the method of optimizing the loading of a train as claimed in any one of claims 1 to 2 when said computer program is executed.
6. A readable storage medium, characterized by: the readable storage medium has stored thereon a computer program which, when executed by a processor, carries out the steps of the method for optimizing the loading of a train as claimed in any one of claims 1 to 2.
CN202210851160.1A 2022-07-20 2022-07-20 Method, device and equipment for optimizing train loading and readable storage medium Active CN115081119B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210851160.1A CN115081119B (en) 2022-07-20 2022-07-20 Method, device and equipment for optimizing train loading and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210851160.1A CN115081119B (en) 2022-07-20 2022-07-20 Method, device and equipment for optimizing train loading and readable storage medium

Publications (2)

Publication Number Publication Date
CN115081119A CN115081119A (en) 2022-09-20
CN115081119B true CN115081119B (en) 2022-11-08

Family

ID=83259676

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210851160.1A Active CN115081119B (en) 2022-07-20 2022-07-20 Method, device and equipment for optimizing train loading and readable storage medium

Country Status (1)

Country Link
CN (1) CN115081119B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115660540B (en) * 2022-12-27 2024-01-26 民航成都物流技术有限公司 Cargo tracking method, cargo tracking device, computer equipment and storage medium
CN117288206B (en) * 2023-11-23 2024-02-02 四川国蓝中天环境科技集团有限公司 Unmanned aerial vehicle route planning method based on self-adaptive large neighborhood search

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455841A (en) * 2013-07-17 2013-12-18 大连海事大学 Container loading method based on improved ant colony algorithm and heuristic algorithm
CN107622321A (en) * 2017-07-27 2018-01-23 山东储备物资管理局八三二处 A kind of algorithm that casing loading pattern is intelligently generated based on multi-constraint condition
CN108520327A (en) * 2018-04-19 2018-09-11 安吉汽车物流股份有限公司 The stowage and device of vehicle-mounted cargo, computer-readable medium
CN110705805A (en) * 2019-10-15 2020-01-17 秒针信息技术有限公司 Cargo assembling method and device, storage medium and electronic device
CN113316116A (en) * 2021-05-27 2021-08-27 南京邮电大学 Vehicle calculation task unloading method based on multi-arm gambling machine
CN114169573A (en) * 2021-11-10 2022-03-11 中铁第四勘察设计院集团有限公司 Method, device and equipment for packing articles and readable storage medium

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20210136608A (en) * 2020-05-08 2021-11-17 삼성에스디에스 주식회사 Apparatus for determining arrangement of objects in space and method thereof
US20220156693A1 (en) * 2020-11-17 2022-05-19 Exel Inc. d/b/a DHL Supply Chain (USA) Computerized system and method for developing optimized cargo transportation solutions

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455841A (en) * 2013-07-17 2013-12-18 大连海事大学 Container loading method based on improved ant colony algorithm and heuristic algorithm
CN107622321A (en) * 2017-07-27 2018-01-23 山东储备物资管理局八三二处 A kind of algorithm that casing loading pattern is intelligently generated based on multi-constraint condition
CN108520327A (en) * 2018-04-19 2018-09-11 安吉汽车物流股份有限公司 The stowage and device of vehicle-mounted cargo, computer-readable medium
CN110705805A (en) * 2019-10-15 2020-01-17 秒针信息技术有限公司 Cargo assembling method and device, storage medium and electronic device
CN113316116A (en) * 2021-05-27 2021-08-27 南京邮电大学 Vehicle calculation task unloading method based on multi-arm gambling machine
CN114169573A (en) * 2021-11-10 2022-03-11 中铁第四勘察设计院集团有限公司 Method, device and equipment for packing articles and readable storage medium

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
Approximation Algorithms for Correlated Knapsacks and Non-martingale Bandits;A. Gupta,R. Krishnaswamy;《2011 IEEE 52nd Annual Symposium on Foundations of Computer Science》;20111222;全文 *
关于带冲突装箱问题的若干优化算法研究;宋园春;《中国优秀硕士学位论文全文数据库 基础学科辑》;20160531;全文 *
基于图着色模型的冲突装箱问题启发式算法;元野等;《运筹与管理》;20130531;第22卷(第5期);全文 *
多车多件货物平衡装载优化研究;朱向等;《工业工程》;20200615(第03期);全文 *
孙宝凤等.考虑货物冲突关系的二维装箱问题研究.《宁波大学学报（理工版）》.2020,第33卷(第2期), *
带冲突关系装箱问题的启发式求解算法;元野，李一军;《运筹与管理》;20150430;第24卷(第2期);第1部分 *
考虑货物冲突关系的二维装箱问题研究;孙宝凤等;《宁波大学学报（理工版）》;20200331;第33卷(第2期);第1部分 *
铁路多车多件货物装载方案优化研究;王海明;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20200331;全文 *

Also Published As

Publication number Publication date
CN115081119A (en) 2022-09-20

Similar Documents

Publication Publication Date Title
CN115081119B (en) Method, device and equipment for optimizing train loading and readable storage medium
Pasha et al. Exact and metaheuristic algorithms for the vehicle routing problem with a factory-in-a-box in multi-objective settings
CN111860754B (en) AGV scheduling method based on ant colony and genetic algorithm
Vahdani Assignment and scheduling trucks in cross-docking system with energy consumption consideration and trucks queuing
CN111553626A (en) Path planning method and device, electronic equipment and storage medium thereof
Kallestad et al. A general deep reinforcement learning hyperheuristic framework for solving combinatorial optimization problems
CN107230028A (en) Vehicle path planning method and device
CN114169573A (en) Method, device and equipment for packing articles and readable storage medium
KR102655411B1 (en) Cold chain delivery route confirmation method, devices, servers and storage media
CN115345549B (en) Vehicle path adjustment method and system combined with loading scheme
Jawarneh et al. Sequential insertion heuristic with adaptive bee colony optimisation algorithm for vehicle routing problem with time windows
Hu et al. Vehicle Routing Problem for Fashion Supply Chains with Cross‐Docking
CN113687651B (en) Path planning method and device for dispatching vehicles on demand
CN110533279B (en) Method, system and storage medium for scheduling cloud logistics platform
Cai et al. Variable neighborhood search for a new practical dynamic pickup and delivery problem
WO2024032376A1 (en) Vehicle path optimization method based on hybrid genetic algorithm, and application thereof
Al-Ashhab et al. Perishable products supply chain network design with sustainability
Alinaghian et al. A novel mathematical model for cross dock open-close vehicle routing problem with splitting
CN109598443B (en) Mission planning method and machine-readable storage medium for vehicle in dynamic environment
CN114970526A (en) Continuous learning method and system for text intention understanding
CN112836846B (en) Multi-depot and multi-direction combined transportation scheduling double-layer optimization algorithm for cigarette delivery
Chen et al. Vehicle Routing Optimization Problem with Time-windows and its Solution by Genetic Algorithm.
CN114862435A (en) Customer loss early warning method, equipment and medium
CN111641974A (en) Method and storage device based on 5G small-sized cellular hybrid renewable energy network
Cai et al. A Two-Layers Heuristic Search Algorithm for Milk Run with a New PDPTW Model

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant