Container port berth and yard allocation joint planning operation method and system

Abstract

The present invention discloses a cluster-based strategy-based joint planning method and system for container port berth and yard allocation. The method comprises: collecting information on ships, container trucks, and yards during a target time period, including the planned arrival time of ships, the number and type of containers on board, the stockpiling status of the yard, and the planned arrival time of external container trucks; inputting the collected information into a pre-established multi-objective mixed integer programming model to calculate an optimal planning strategy, including a berthing plan for all ships and a loading and unloading plan for container batches during the specified time period; wherein the pre-established multi-objective mixed integer programming model includes three objective functions, namely, minimizing the expected turnaround time for ship berthing, minimizing the total container transport distance between berths and container areas, and minimizing the workload imbalance in the container areas of the yard; and executing ship and yard operations according to the optimal planning strategy. The present invention can improve the overall operational efficiency of the terminal.

Description

Technical Field

[0001] The present invention relates to the technical field of container terminal operation resource allocation, and in particular to a cluster strategy-based container port berth and yard allocation joint planning method and system. Background Art

[0002] With the booming global trade, maritime transport plays a vital role in the global economic and trade system. According to statistics, over 80% of global cargo relies on ships. In recent years, global container trade has continued to grow, and some Asian ports have even experienced severe congestion. Against this backdrop, improving the efficiency of container terminal operations has become a key requirement for the industry's development. This efficiency, in turn, depends largely on the smooth operation of container terminals and the effective utilization of terminal resources. Berths and storage yards are the most critical and limited resources at terminals, and their rational allocation is crucial to the efficient operation of the entire terminal.

[0003] Currently, most ports generally adopt a consignment strategy to manage container areas in their yards. This involves dividing the container area into several sub-blocks, each with a fixed number of contiguous bays, and then storing containers in these sub-blocks. However, due to the uncertainty of container arrival and departure times, certain container areas may be overloaded or overcrowded during certain time periods, while other areas remain idle for long periods of time. This unbalanced workload distribution prevents optimal resource allocation and seriously restricts the improvement of overall terminal operational efficiency. Summary of the Invention

[0004] The main purpose of the present invention is to provide a method and system for joint planning of container port berth and yard allocation, which can improve the overall operational efficiency of the terminal.

[0005] The technical solution adopted in the present invention is:

[0006] A method for joint planning of container port berth and storage yard allocation based on cluster strategy is provided, comprising the following steps:

[0007] S1. Collect information on ships, container trucks, and container yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board, the storage status of the container yard, and the planned arrival time of the container trucks.

[0008] S2. Input the collected information into a pre-established multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the assigned container yard area, loading and unloading routes, and loading and unloading time for the corresponding container batch on the ship or truck. The pre-established multi-objective mixed integer programming model includes three objective functions, namely, minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and container yard areas, and minimizing the workload imbalance in the container yard areas.

[0009] S3. Execute ship and yard operations according to the optimal planning strategy.

[0010] Following the above technical solution, in step S1, the containers on the same ship or container truck are divided into multiple batches according to their destinations, and then the batches with the same destination are grouped into a container cluster. The container cluster is assigned to a container area template using an integer coding scheme, and then the container area template is assigned to a specific container area in the yard; the ship or container truck itself is used as the destination.

[0011] Following the above technical solution, when solving a multi-objective mixed integer programming model, the optimal solution is selected based on the fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy.

[0012] Following the above technical solution, in the process of screening the optimal solution based on the fuzzy correlation entropy fitness evaluation mechanism FCE-FEM, the global search capability is enhanced through an improved taboo search strategy. Specifically, based on the taboo search and multi-neighborhood search structure, a variety of neighborhood movement rules are designed to increase the neighborhood solution set, and the mutated neighborhood solution set is re-optimized.

[0013] Following the above technical solution, multiple neighborhood movement rules include the Exchange rule, the Fragment Insert rule, and the Intermingle Swap rule. The Exchange rule randomly selects two container slots and exchanges the genes of the two container areas. The Fragment Insert rule first randomly selects two container slots, removes the genes between the two container slot numbers, and then randomly selects a position in the remaining chromosomes to insert the removed genes into the position. The Intermingle Swap rule randomly generates several container area positions and disrupts the gene order at the corresponding ship positions.

[0014] Following the above technical solution, the constraints of the multi-objective mixed integer programming model include: each ship can only choose one berth to berth, two ships berthed at the same berth must have a berthing order, each ship can only berth and leave the berth once, the start and end times of loading and unloading of the ship are specified, the ship is required to start operations immediately after berthing, and the ship must leave the berth immediately after completing operations, and the loading and unloading operation time of the container batch is within the loading and unloading time of the corresponding ship.

[0015] Following the above technical solution, when solving the mixed integer programming model, the objective function value corresponding to each planning strategy is calculated in a specific order based on the three constructed objective function formulas. First, the ship's berthing time is added to the known port time to accurately calculate the ship's departure time, and then the cost incurred by the ship's deviation from the expected turnaround time interval is calculated according to the relevant formula; secondly, within the ship's berthing time range, the loading and unloading operation plan of the corresponding container batch is generated, and the workload imbalance within the planning range is calculated; finally, based on the loading and unloading operation plan and the ship's berthing position information, the storage location of each container batch in the yard is clarified, and the average transportation distance of the container batch is calculated. Then, the average moving distance of all container batches is accumulated according to the formula, and the total container movement cost is finally obtained.

[0016] The present invention also provides a cluster strategy-based container port berth and yard allocation joint planning operation system, comprising:

[0017] The information collection module is used to collect information about ships, container trucks, and storage yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board the ship, the storage status of the storage yard, and the planned arrival time of the container trucks;

[0018] The optimal planning strategy solving module is used to input the collected information into a pre-built multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the yard container area, loading and unloading route, and loading and unloading time allocated to the container batch on the corresponding ship or container truck. The pre-built multi-objective mixed integer programming model includes three objective functions: minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and yard container areas, and minimizing the workload imbalance in the yard container area.

[0019] Actuators for executing vessel and yard operations according to optimal planning strategies.

[0020] Following the above technical solution, when solving a multi-objective mixed integer programming model, the optimal planning strategy solving module specifically selects the optimal solution based on the fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy.

[0021] The present invention also provides a computer storage medium storing a computer program executable by a processor, wherein the computer program executes the cluster strategy-based container port berth and yard allocation joint planning operation method described in the above technical solution.

[0022] The beneficial effects of the present invention are as follows: the present invention constructs a multi-objective mixed integer programming model, wherein the multi-objectives are respectively to minimize the deviation of the ship berthing from the expected turnaround time, to minimize the total container transportation distance between the berth and the container area of ​​the yard, and to minimize the imbalance of the container area workload of the container area of ​​the yard. The three objectives are comprehensively considered from three different angles: port berthing planning, yard transportation route, and yard workload. The three objectives are closely linked. The first objective generates a ship berthing plan, clarifies the berthing position and berthing time, and aims to minimize the deviation from the expected turnaround time of the ship. Based on the first objective, various container loading and unloading routes can be determined to obtain a loading and unloading plan with the minimum transportation distance to achieve the second objective. Based on the second objective, whether the workload of the container area of ​​the yard is balanced is calculated, and the most balanced workload solution is found to achieve the third objective. The present invention solves complex port operations through a mathematical model, and can efficiently obtain a container port berth and yard space allocation plan for a future time period, which is conducive to improving the utilization rate of the yard space and reducing congestion on the yard side.

[0023] Furthermore, in the process of screening the optimal solution using the fuzzy correlation entropy-based fitness evaluation mechanism FCE-FEM, three new neighborhood movement rules were designed through an improved taboo search to increase the neighborhood solution set, and the mutated neighborhood solution set was re-optimized, thereby improving the global search capability of the algorithm.

[0024] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 This is a flow chart of a method for joint planning of container port berth and storage yard allocation based on cluster strategy according to an embodiment of the present invention;

[0027] Figure 2 This is a schematic diagram of the layout of a container terminal according to an embodiment of the present invention;

[0028] Figure 3 This is a flow chart of an adaptive multi-objective evolutionary algorithm according to an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of a container cluster entering and exiting a container yard area according to an embodiment of the present invention;

[0030] Figure 5 This is a design diagram of an improved tabu search neighborhood movement rule according to an embodiment of the present invention;

[0031] Figure 6A Schematic diagram of the objective function value of the ship berthing deviation time under different berth throughput capacities according to one embodiment of the present invention;

[0032] Figure 6B Schematic diagram of objective function values ​​for workload imbalance under different berth throughput capacities according to an embodiment of the present invention;

[0033] Figure 6C Schematic diagram of the objective function value of container transportation distance under different berth throughput capacities according to one embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0035] It should be noted that the illustrations provided in the embodiments of the present invention are only schematic illustrations of the basic concept of the present invention. Therefore, the drawings only show components related to the present invention and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0036] In the present invention, it should also be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer" and the like are used to indicate positions or locations based on those shown in the accompanying drawings. These terms are intended solely to facilitate the description of the present application and to simplify the description. They are not intended to indicate or imply that the devices or components referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present application. Furthermore, the terms "first" and "second" are used solely for descriptive and distinguishing purposes and should not be construed as indicating or implying relative importance.

[0037] In addition, it should be noted that the features of the various embodiments of the present invention may be combined or coupled in part or in whole, and, as will be appreciated by those skilled in the art, may interact and operate in different ways. Each embodiment may be implemented independently of one another or in an associated relationship.

[0038] Example 1

[0039] like Figure 1 As shown, the container port berth and yard allocation joint planning operation method based on cluster strategy in an embodiment of the present invention includes the following steps:

[0040] S1. Collect information on ships, container trucks, and container yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board, the storage status of the container yard, and the planned arrival time of the container trucks.

[0041] S2. Input the collected information into a pre-established multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the assigned container yard area, loading and unloading routes, and loading and unloading time for the corresponding container batch on the ship or truck. The pre-established multi-objective mixed integer programming model includes three objective functions, namely, minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and container yard areas, and minimizing the workload imbalance in the container yard areas.

[0042] S3. Execute ship and yard operations according to the optimal planning strategy.

[0043] Furthermore, in step S1, containers on the same ship or truck are divided into multiple batches based on their destinations. Batches with the same destination are then grouped into a container cluster. An integer encoding scheme is used to assign the container cluster to a container area template, which is then assigned to a specific container area in the yard. The ship or truck itself serves as the destination. A container area template can be considered a simple copy of a container area in the yard. A container area template is a combination of clusters consisting of different container batches. These clusters are stacked together within the specified range to form a container area template, reducing the total workload difference between container area templates (yard container areas) within the specified range. Finally, each container area template is assigned a container area in the yard to minimize the travel distance of all container clusters.

[0044] When solving multi-objective mixed integer programming models, the fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy is used to screen the optimal solution.

[0045] In the process of screening the optimal solution using the fuzzy correlation entropy-based fitness evaluation mechanism FCE-FEM, the global search capability is enhanced through an improved tabu search strategy. Specifically, based on the tabu search and multi-neighborhood search structure, a variety of neighborhood movement rules are designed to increase the neighborhood solution set, and the mutated neighborhood solution set is re-optimized.

[0046] Among them, the various neighborhood movement rules include the Exchange rule, the Fragment Insert rule, and the Intermingle Swap rule. The Exchange rule randomly selects two container slots and exchanges the genes of the two container areas; the Fragment Insert rule first randomly selects two container slots, removes the genes between the two container slot numbers, and then randomly selects a position in the remaining chromosomes to insert the removed genes into the position; the Intermingle swap rule randomly generates several container area positions and disrupts the gene order at the corresponding ship positions.

[0047] Furthermore, the constraints of the multi-objective mixed integer programming model include: each ship can only choose one berth to berth, two ships berthed at the same berth must have a berthing order, each ship can only berth and leave the berth once, the start and end times of loading and unloading of the ship are specified, the ship is required to start operations immediately after berthing, and leave the berth immediately after finishing operations, and the loading and unloading operation time of the container batch is within the loading and unloading time of the corresponding ship.

[0048] Preferably, when solving the mixed integer programming model, the objective function value corresponding to each planning strategy is calculated in a specific order according to the three constructed objective function formulas. First, the ship's berthing time is added to the known port time to accurately calculate the ship's departure time, and then the cost incurred by the ship's deviation from the expected turnaround time interval is calculated according to the relevant formula; secondly, within the ship's berthing time range, the loading and unloading operation plan of the corresponding container batch is generated, and the workload imbalance within the planning range is calculated; finally, based on the loading and unloading operation plan and the ship's berthing position information, the storage location of each container batch in the yard is clarified, and the average transportation distance of the container batch is calculated, and then the average moving distance of all container batches is accumulated according to the formula to finally obtain the total container movement cost.

[0049] In addition to using the fuzzy correlation entropy-based fitness evaluation mechanism (FCE-FEM) to select optimal solutions, other fitness calculation methods can also be applied to select better solutions. For example, in the NSGA-II algorithm (multi-objective genetic algorithm), non-dominated sorting and crowding calculation are used to maintain population diversity and gradually approach the optimal solution set. The final solution is typically selected after the algorithm completes, by selecting one or more solutions from the optimal solution set of the last generation based on the decision maker's preferences or specific rules. Non-dominated sorting refers to a solution that is not superior to another solution in all objective functions and is strictly inferior to another solution in at least one objective function. For example, when solving three minimization objective functions, suppose the three objective functions of solution A are (60, 70, 85) and the three objective functions of solution B are (65, 75, 86). If all three solutions in A are superior to the three solutions in B, then A is said to dominate B, making A a non-dominated solution and B a dominated solution. This process continues in this way. After several iterations, the algorithm will find a set of non-dominated solutions, i.e., the optimal solution set. These are all optimal solutions.

[0050] This invention constructs a multi-objective mixed integer programming model, specifically minimizing the expected turnaround time for ship berthing, minimizing the total container transport distance between berths and container yards, and minimizing the imbalance in container yard workload. These three objectives are comprehensively considered from the perspectives of port berthing planning, container yard transport routes, and container yard workload, and are closely linked. By solving complex port operations through mathematical modeling, this invention can efficiently determine a plan for allocating container port berths and container yard space for a future time period, thereby improving container yard space utilization and reducing congestion on the container yard side.

[0051] Example 2

[0052] This example is a specific optimization based on Example 1. A mixed integer programming model with multiple objectives is pre-established, with the goal of minimizing the deviation of ship berthing from the expected turnaround time, the total container transport distance between berths and container areas, and the imbalance of container area workload. This model considers constraints such as container storage area restrictions, cluster allocation uniqueness, container destination consistency within a cluster, container area workload balance, a total number of bays in a container area, and a limit on the number of bays in a cluster. To solve this complex model, a cluster stacking strategy can be used to allocate yard space. A cluster refers to a number of consecutive bays in a container area. Containers are divided into batches based on the number of unloading and loading vessels (trucks). Each vessel is considered a departure or destination, and trucks arriving at the port at the same time are also considered a departure or destination. Therefore, containers can be divided into clusters based on their destination. Container batches with the same destination form a cluster and are stacked in the corresponding container area. The concept of yard container area template was designed. The container area template can be regarded as a simple copy of the yard container area. The container area template is a combination of clusters consisting of different container batches. These clusters are stacked together within the specification range to form a container area template to reduce the total workload gap between container area templates (yard container areas) within the specification range, and a yard container area is assigned to each container area template to minimize the moving distance of all container clusters.

[0053] In this embodiment, an adaptive multi-objective evolutionary algorithm is designed to solve the mixed integer programming model. In the algorithm, a fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy is proposed to evaluate the multi-objective method, and an adaptive local enhancement search strategy is introduced to improve the global search capability.

[0054] After the above model and solution algorithm are determined, data is collected based on the model's parameters and then fed into the model for solution, resulting in the optimal planning strategy within a specified timeframe. The collected data includes information such as the vessel's estimated arrival time, the number and type of containers on board, the yard's stockpiling status, and the arrival time of external trucks. Container batches are divided into multiple container clusters based on the originating and destination vessels or trucks. Yard space can be allocated using a cluster stacking strategy. A cluster refers to a number of consecutive, adjacent bays within a yard's container area. Containers are divided into batches based on the unloading and loading vessels (trucks). Each ship is considered a departure or destination, and trucks arriving at the port during the same time period are considered a departure or destination. Therefore, container clusters can be divided based on the destination of the containers. Container batches with the same destination form a cluster and are stacked in the corresponding container area of ​​the yard.

[0055] Specifically, this embodiment provides a multi-objective mixed integer programming model for the joint planning problem of container port berth and yard allocation based on clustering strategy, which is expressed as:

[0056]

[0057]

[0058]

[0059]

[0060] Where the objective function is to simultaneously minimize the deviation of the ship from the expected turnaround time, the total distance moved by the container, and the workload imbalance in the container yard area. Indicates that the ship's berthing deviates from the expected turnaround time, Indicates the total transport distance of the container group between the berth and the container area of ​​the yard. It indicates that the workload in the container area of ​​the yard is unbalanced. Ships Actual berthing time; Ships The actual departure time; For ships the expected turnaround service interval; express The maximum workload processed in all bins during the time period; express The minimum amount of work processed in all bins during the time period; If the vessel to be unloaded Docked at the berth If yes, it is 1, otherwise it is 0; If the ship to be loaded with boxes Docked at the berth If yes, it is 1, otherwise it is 0; Indicates that if there is a group of containers from the ship Unloading and finally loading on board ship 1 if on, 0 otherwise; Indicates that if the container batch Belong to the cluster 1 when it is, otherwise 0; Indicates Time period box template Assign to cluster The number of bays; Represents the box area template Assigned to box area If yes, it is 1, otherwise it is 0; Indicates box area Distance to berth The average distance between . Represents a set of time periods within the problem cycle; represents a collection of berths; Represents a collection of ships; Indicates the collection of ships (container trucks) that need to unload containers; Indicates the collection of ships (container trucks) that need to be loaded; Represents a collection of bins; Represents a collection of clusters; Represents a collection of bin templates.

[0061] To decompose a highly complex optimization model, first, The formula (f1) is used to generate a ship's berthing plan, specifying berthing locations and times, with the goal of minimizing deviations from the ship's expected turnaround time. Secondly, based on the actual ship arrival and departure times determined by f1, different container batches are stacked according to their destinations in different yard container templates. A yard container template can be considered a simple copy of a yard container template. As the yard container templates are stacked, the workload imbalance of the yard container templates is minimized using formula f3. The workload of a yard container template in each time period is the sum of the number of containers entering and leaving the container template during that time period. For example, if in the second time period, container template k has three bays of containers entering and four bays of containers leaving, the workload of yard template k in that time period is 7. Similarly, the workload of other yard container templates is calculated to obtain the workload imbalance of the yard container templates within that time period (maximum minus minimum). Finally, each container template is assigned a yard container, and the movement distance of all container clusters is minimized using formula f2.

[0062] Furthermore, the constraints of the multi-objective mixed integer programming model include:

[0063] 1) Each vessel can only choose one berth to berth

[0064]

[0065] in, is a 0-1 variable. If the berth Assigned to ships If yes, it is 1, otherwise it is 0.

[0066] 2) Two ships berthed at the same berth must have a berthing order of front and back

[0067]

[0068]

[0069] A 0-1 variable, if the berth Assigned to a ship and ships , and ships On ships Previously docked at the berth 1 if yes; 0 otherwise.

[0070] 3) Each vessel can only berth and unberth once

[0071]

[0072]

[0073] is a 0-1 variable, indicating that During this period, the ship If it is docked, it is 1, otherwise it is 0;

[0074] is a 0-1 variable, indicating that During this period, the ship It is 1 when leaving the berth, 0 otherwise.

[0075] 4) Specify the start and end time of loading and unloading of ships

[0076]

[0077]

[0078] is a 0-1 variable, indicating that Ships within the time period It is 1 when loading and unloading operation starts, otherwise it is 0;

[0079] is a 0-1 variable, indicating that Ships within the time period It is 1 when the loading and unloading operation is completed, otherwise it is 0.

[0080] 5) Requires that operations begin immediately after a vessel berths and that the vessel leave the berth immediately after completing operations

[0081]

[0082]

[0083] 6) Indicates the constraints for starting and ending loading and unloading of ships

[0084]

[0085]

[0086] in, Indicates ship Time of commencement of loading and unloading operations; Indicates ship The time when loading and unloading operations are completed.

[0087] 7) Relationship between vessel operation start time and operation end time

[0088]

[0089] in, Indicates ship Time of loading and unloading operations.

[0090] 8) For the same bay, the relationship between the operation time of the two ships in front and behind

[0091]

[0092] in, Represents a large positive constant.

[0093] 9) The ship's start and end operation time must be within the feasible service time window

[0094]

[0095]

[0096] in, Indicates ship feasible service time period.

[0097] 10) The amount of dock work generated by loading and unloading containers must be subject to the berth's handling capacity.

[0098]

[0099] in, is a 0-1 variable. If the container batch If it is a transfer box, it is 1, otherwise it is 0; is a 0-1 variable. If the container batch If it is an export box, it is 1, otherwise it is 0; is a 0-1 variable. If the container batch 1 if it is an import box, 0 otherwise. It represents the maximum workload that berth b can handle in a unit time period.

[0100] 11) Indicates that the workload of each container batch should be equal to its task volume

[0101]

[0102]

[0103] in, Indicates Time period will be container batch The workload of unloading from the ship to the storage yard; Indicates Time period will be container batch The workload of loading from the yard to the ship.

[0104] 12) Indicates that the operation time of the container batch should be within the loading and unloading time of the corresponding ship

[0105]

[0106]

[0107] in, is a 0-1 variable. During the time period, if the container batch If the ship is unloaded to the yard, it is 1, otherwise it is 0; is a 0-1 variable. During the time period, if the container batch It is 1 when loading from the yard to the ship, otherwise it is 0.

[0108] 13) Export boxes and transit boxes are placed in the export box area, and import boxes are placed in the import box area

[0109]

[0110]

[0111] in, If the cluster Assign to bin template If yes, it is 1, otherwise it is 0; Represents the set of export box areas; Represents a collection of import bins.

[0112] 14) Ensure that all containers in the cluster have storage locations and that the storage locations freed up by the containers loaded onto the ship are equal to the locations released by the corresponding cluster

[0113]

[0114]

[0115] in, Indicates Time period due to cluster After being released, the box area template The number of extra bays in the middle.

[0116] 15) Indicates that a container batch can only be assigned to one cluster

[0117]

[0118] 16) A container area template can only be assigned to one container area in a yard. Similarly, a container area in a yard can only select one container area template.

[0119]

[0120]

[0121] 17) Containers with the same destination must be stored in the same cluster. Containers stored in any cluster can only have one destination.

[0122]

[0123]

[0124] in, is a 0-1 variable. If the cluster Reserved for loading onto ships (container trucks) The value is 1 if the container batch is on, and 0 otherwise.

[0125] 18) Indicates each box area Balanced workload within a time period

[0126]

[0127]

[0128] in, Indicates container batch The total number of bays required; It is a 0-1 variable. If the export box batch exist If the server arrives at the storage yard within the time period, the value is 1; otherwise, the value is 0. Is a 0-1 variable, if the import box batch exist If the container arrives at the storage yard within the specified time, the value is 1; otherwise, the value is 0. Indicates that it is located in the container area of ​​the yard The collection of clusters in .

[0129] 19) Indicates that the number of bays occupied by clusters of the same box template cannot exceed the total number of bays in the box template

[0130]

[0131] in, Indicates the total number of bays in the container area template of the yard.

[0132] 20) Indicates that the number of container bays allocated to the cluster cannot exceed the maximum number of bays that the cluster can accommodate

[0133]

[0134] in, Represents a cluster The maximum number of bays that can be accommodated.

[0135] 21) Represents decision variable constraints

[0136] Furthermore, the steps of the adaptive multi-objective evolutionary algorithm are as follows:

[0137] 1) Encoding and decoding

[0138] A two-layer integer encoding scheme is used to encode and represent ship berthing plans. The first-layer integer represents the ship's berthing time, while the second-layer integer represents the ship's berthing location. Container batches are also divided into clusters based on their destination information. Clusters are then assigned to container area templates using the integer encoding scheme, and the container area templates are then assigned to specific container areas in the yard.

[0139] During the decoding process, the objective function values ​​corresponding to each coding scheme are calculated sequentially according to the objective function formula of the established multi-objective mixed integer programming model. First, the ship's berthing time is added to the known port time to accurately calculate the ship's departure time. The cost of the ship's deviation from the expected turnaround time is then calculated using the relevant formula. Second, based on the allocation plan of clusters and container area templates, the cluster set to which the container area template belongs is determined in detail. Within the ship's berthing time range, a loading and unloading operation plan for the corresponding container batch is generated. Based on this, the loading and unloading operation plan for the container area template and the comprehensive loading and unloading operation plan for all container area templates are further derived. The container area workload imbalance within the planning range is calculated using the corresponding formula. Finally, based on the allocation plan between container area templates and container areas and the ship berthing location information, the storage location of each cluster and container batch in the yard is determined. The average transportation distance between the berth and the container area for each container batch is calculated. The average movement distance of all container batches is then accumulated according to the formula to finally obtain the objective function value of the total container movement cost.

[0140] 2) Fitness evaluation mechanism based on fuzzy correlation entropy (FCE-FEM method)

[0141] Construct comparison points, dynamic reference points, and dynamic worst-case points. The calculation formula for comparison points is determined based on the objective function values ​​of individuals in the population at different iteration stages and is used to measure the relative position of individuals in the target space. The calculation formula for dynamic reference points comprehensively considers factors such as the number of iterations, individual target values, and the total number of targets, providing a dynamic reference standard for evaluating individual performance. The calculation formula for dynamic worst-case points is also based on individual information in the population and is used to determine the worst-case performance of individuals in the target space.

[0142] The relative membership function is used to map the membership values ​​of sub-goals. By introducing upper and lower bound factors, the sub-goal values ​​are converted into membership values, making different goals comparable and facilitating subsequent calculation and evaluation.

[0143] The contrast points and dynamic reference points are converted into contrast fuzzy sets and ideal fuzzy sets, and the fuzzy set theory is used to fuzzify the performance of individuals in the multi-objective space to more accurately describe the individual's quality.

[0144] The fuzzy entropy, fuzzy partial entropy, fuzzy correlation entropy, and fuzzy correlation entropy coefficient are calculated and compared between fuzzy sets and ideal fuzzy sets. Fuzzy entropy measures the uncertainty of a fuzzy set, while fuzzy partial entropy further assesses the degree of deviation of an individual from the ideal point on a specific objective. Fuzzy correlation entropy reflects the similarity between an individual and the ideal individual by calculating the degree of correlation between two fuzzy sets. Finally, by calculating the fuzzy correlation entropy coefficient (FCE coefficient), the multi-objective problem is transformed into a simple and intuitive coefficient used to evaluate the performance of the solution. A higher coefficient indicates a better solution.

[0145] Among them, the fuzzy correlation entropy fitness evaluation mechanism (FCE-FEM) has the following specific steps:

[0146] Step 1: Construct comparison points , dynamic reference point and the worst dynamic point .

[0147]

[0148]

[0149]

[0150] Where: Indicates in Individuals in the iteration No. goals, , Refers to individuals in a population, , Refers to the maximum number of iterations; and Respectively represent The target in the iteration The best and worst values ​​of , Refers to the total number of targets.

[0151] Step 2: Use relative membership function to map sub-goals The membership value of .

[0152]

[0153] Where: , , and Represent the upper and lower bound factors ( ).

[0154] Step 3: Convert the contrast points and dynamic reference points into contrast fuzzy sets and ideal fuzzy sets .

[0155]

[0156]

[0157] Step 4: Calculation and The fuzzy entropy of , .

[0158]

[0159]

[0160] Step 5: Calculation about The fuzzy partial entropy of about The fuzzy partial entropy of .

[0161]

[0162]

[0163] Step 6: Calculation and The fuzzy correlation entropy between them.

[0164]

[0165] Step 7: Calculation and The fuzzy correlation entropy coefficient between express.

[0166]

[0167] Where:

[0168] 3) Evolutionary Operator

[0169] Selection based on the FCE coefficient: Using a roulette wheel, the calculated FCE coefficient is used to determine the selection probability and cumulative probability of an individual. The selection probability formula is based on the ratio of an individual's FCE coefficient to the sum of the FCE coefficients of all individuals in the population, while the cumulative probability is calculated by adding up the individual selection probabilities. This method favors individuals with higher fitness (i.e., larger FCE coefficients) during selection, ensuring that the population evolves towards the optimal solution.

[0170] Collaborative single-point crossover: Collaborative single-point crossover operations are performed sequentially for berth planning, allocation between cluster and bin modules, and allocation between bin templates and bins. The crossover probability uses an adaptive adjustment mechanism, and its calculation formula comprehensively considers factors such as the upper and lower bounds of the crossover probability, the individual fuzzy correlation entropy coefficient, the maximum fuzzy correlation entropy coefficient in the population, and the average correlation entropy coefficient. During the crossover operation, a real number is first randomly generated and compared with the crossover probability. If it is less than the crossover probability, two parent chromosomes are randomly selected from the population after the FCE-based selection operation. Then, a crossover point is randomly generated, and the genes of the specified parts of the parent chromosomes are exchanged to form two new chromosomes. Finally, the genes of the newly generated chromosomes are checked for duplicate bin numbers and corrected if any to ensure the feasibility of the new chromosomes.

[0171] 2-Opt Reversal Mutation: 2-Opt reversal mutation operations are performed sequentially on the three types of chromosomes mentioned above. The mutation probability also uses an adaptive adjustment mechanism, and its calculation formula is related to the upper and lower bounds of the mutation probability. During the mutation operation, a random real number is generated. If it is less than the mutation probability, a chromosome is randomly selected from the population. Two random bins are then generated, and the gene segment between these two bins is reversed to form the mutated offspring chromosome. This mutation method significantly perturbs the parent chromosome, helping to increase population diversity and prevent premature convergence of the algorithm.

[0172] 4) Adaptive local reinforcement search strategy

[0173] An adaptive strategy based on pseudo-entropy theory: The FCE coefficient is integrated into pseudo-entropy theory, and the pseudo-entropy value for multi-objective problems is calculated using a specific formula. The pseudo-entropy value takes into account factors such as the FCE coefficient of each individual in the population, the total number of objectives, and the number of iterations, and is used to measure the distribution of the population in the multi-objective space. The pseudo-entropy value determines whether to initiate local search. When the pseudo-entropy value meets certain conditions, indicating low population diversity and the possibility of being trapped in a local optimum, local search is initiated to improve the algorithm's ability to escape the local optimum.

[0174] Improved Tabu Search: Based on the tabu search and multi-neighborhood search architecture, three neighborhood movement rules are designed: Exchange, Fragment Insert, and Intermingle Swap. The Exchange rule randomly selects two bins and swaps the genes between them. The Fragment Insert rule first randomly selects two bins, removes the genes between them, and then randomly selects a position on the remaining chromosome to insert the removed genes into that position. The Intermingle Swap rule randomly generates several bins and shuffles the gene order at the corresponding bin positions. When executing the improved tabu search, parameters such as the neighborhood structure, maximum number of iterations, number of iterations, tabu length, tabu table, and number of neighborhood solutions are first defined. An individual is then randomly selected from the current solution set. A new neighborhood solution set is generated for this individual according to the neighborhood rules, and its fitness is calculated. Find the individual with the highest fitness value in the neighborhood solution set. If it meets certain conditions (such as not being in the taboo table and having a fitness value better than the current optimal solution), update the current optimal solution and the taboo table. Otherwise, select the individual with the best fitness value from the neighborhood solution set that is not in the taboo table as the new current solution, and update the taboo table. Repeat the above process until the maximum number of iterations is reached, and output the optimal solution. This will further optimize the mutated solution and improve the algorithm's global search capability.

[0175] Specifically, the adaptive local reinforcement search strategy includes:

[0176] 1) Adaptive strategy based on pseudo-entropy theory

[0177] The fitness value coefficient is integrated into the pseudo-entropy theory to calculate the pseudo-entropy value of the multi-objective problem, and whether to start local enhanced search is determined based on the pseudo-entropy value.

[0178]

[0179]

[0180] Where: is the pseudo-entropy value of the g-th iteration.

[0181] 2) Improved tabu search

[0182] Define three neighborhood structures including neighborhood movement rules based on Exchange ( ), Neighborhood movement rules based on FragmentInsert ( ) and the neighborhood movement rule based on Intermingle swap ( ), re-optimize the mutated solution, if , then select the mutated individuals and perform the following operations:

[0183] Step 1: Define the neighborhood structure ; Maximum number of iterations Gen; Iteration number ; Taboo length Taboo table ; Number of neighborhood solutions .

[0184] Step 2: Randomly select an individual from the above solution set ,make .

[0185] Step 3: Targeting Individuals Based on neighborhood rules Generate a new neighborhood solution set , and calculate its fitness value at the same time.

[0186] Step 4: Find the highest fitness value, i.e. , and record the corresponding solution set .

[0187] Step 5: If ,make Update taboo table , go to step 7; otherwise, go to step 6.

[0188] Step 6: Solve the problem from the neighborhood Select the one with the best fitness value And not on the taboo list The solution set of .make, , and update the taboo table .

[0189] Example 3

[0190] This embodiment is a specific implementation of the above method embodiment, wherein data collection and preparation are specifically as follows:

[0191] 1) Instance data generation

[0192] The embodiment of the present disclosure generates a set of test cases based on the actual operation status of a container terminal with an annual throughput of 2 million TEUs. The container terminal layout and container transshipment process are as follows: Figure 2 As shown in the figure, the number of berths is determined to be 5, with 2 import and 8 export container areas, respectively. The number of ships is 10, the number of container lots is 34, and the number of bays per container area is set to 50. The problem period is 18 time periods, each of which is 3 hours. The berth throughput capacity is 25 bays per time period, and the maximum container volume that the cluster can accommodate is 30 bays, with 1 bay containing 50 TEUs. The distance from the berth to the container area is considered the Manhattan distance between the center points. For each container lot, the number of container bays required follows a uniform distribution [5, 10]; the loading and unloading time for each ship also follows a uniform distribution [2, 3]. Each container lot type (import, export, or transit) is randomly generated to ensure that all test cases meet the feasibility requirements of actual operations.

[0193] 2) Algorithm parameter settings

[0194] After multiple parameter experiments, the parameters of the adaptive multi-objective evolutionary algorithm are set as follows: population size is 600, maximum number of iterations is 100, minimum crossover probability is 0.4, maximum crossover probability is 0.99, minimum mutation probability is 0.2, maximum mutation probability is 0.6, upper bound factor and the lower bound factor .

[0195] The multi-objective evolutionary algorithm process is as follows Figure 3 The specific steps are as follows:

[0196] 1) Coding scheme

[0197] First, a two-layer integer encoding scheme is adopted Describes the ship's berthing plan, with the first layer representing the ship's berthing time , the second layer represents the berthing position of the ship ,in and Respectively represent The container batches are divided into several clusters according to the container destination, and then the clusters are assigned to the container area template using an integer coding scheme. and assigning container area templates to yard container areas ,in Indicates the number of the f-th cluster allocation box template, Indicates the number of the yard container area assigned to the k-th container area template.

[0198] 2) Decoding process

[0199] During decoding, the objective function value of each scheme is calculated based on the encoding information and model calculation rules.

[0200] Calculate the deviation of the ship from the expected turnaround time: For each ship, its berthing time is , the departure time is The ship deviates from the expected turnaround time interval by ,in and are the left and right boundaries of the expected turnaround service time of the ship respectively.

[0201] Calculate the total container transport distance: Based on the cluster and container area template allocation plan, determine the loading and unloading location of each container batch, calculate the transportation distance from the berth to the container area in the yard, and from the container area in the yard to the berth, and accumulate the total movement distance of all container batches, including the movement distance of transit containers, export containers, and import containers.

[0202] Calculating workload imbalance in the container area: Consider the time and workload distribution of containers entering and leaving the yard under the clustering strategy. Since the cluster reserves space for the container during its stay, once the container is removed, the cluster will be released. Therefore, when calculating the workload, it is necessary to accurately count the number of loading and unloading operations in the container area in each time period based on the arrival and departure times of the containers in the cluster, such as Figure 4 For a container area containing multiple clusters, when calculating workload imbalance, it is necessary to consider the loading and unloading conditions of containers in each cluster and the differences in operating time between different clusters to ensure accurate calculation of the workload of the container area in different time periods, thereby obtaining an accurate workload imbalance value.

[0203] 3) Fitness evaluation mechanism based on fuzzy correlation entropy (FCE-FEM method)

[0204] For each individual in the population, the fitness value is calculated according to the FCE-FEM method of the above embodiment.

[0205] 4) Select an operation

[0206] Based on the calculated FCE coefficient, the roulette wheel rule is used for selection. The probability of selection and cumulative probability By randomly generating a real number ,like , then select individual Enter the next generation of population and repeat the process until the appropriate number of individuals are selected.

[0207] 5) Collaborative single-point cross operation

[0208] Perform a collaborative single-point crossover operation on the selected individuals. Take the allocation scheme between the bin template and the bin as an example. First, randomly generate a real number , and the crossover probability For comparison, the crossover probability is obtained using an adaptive adjustment mechanism. , randomly select two parent chromosomes from the population after the selection operation and . Randomly generate an intersection point , exchange parent chromosomes and in arrive columns, forming two new chromosomes and Next, examine the newly generated chromosomes and Is there a duplicate box number in the gene? If so, find the duplicate box number and record it as and , in exchange chromosomes of and in , forming two new feasible chromosomes. Similar crossover operations are also performed on the berth planning scheme and the allocation scheme between cluster and box area modules.

[0209] 6) 2-Opt reverse mutation operation

[0210] Taking the bin template and the allocation scheme between bins as an example, a real number is randomly generated If the mutation probability , where the mutation probability is calculated according to the adaptive adjustment mechanism, then a chromosome is randomly selected from the population after the selection operation , randomly generate two boxes and , the gene fragment between the two boxes in the chromosome is reversed to form the offspring chromosome Similar mutation operations are performed on the berth plan chromosomes and the cluster and bin template allocation plan chromosomes.

[0211] 7) Adaptive local reinforcement search

[0212] ① Adaptive strategy based on pseudo-entropy theory

[0213] According to the FCE coefficient of individuals in the current population, calculate the The pseudo-entropy value of the iteration . Set a threshold (can be determined based on experience or experiment), if , it indicates that the population diversity is low and may fall into a local optimum. At this time, the improved tabu search is started; otherwise, the next round of iteration is continued.

[0214] ②Improved tabu search

[0215] Taking the allocation plan between container area templates and container areas as an example, the neighborhood structure is defined, including three neighborhood movement rules: Exchange, Fragment Insert, and Intermingle swap. Figure 5 shown.

[0216] 8) Algorithm completion and result output

[0217] Repeat the fitness evaluation, evolutionary operations (selection, crossover, mutation) and adaptive local reinforcement search steps until the maximum number of iterations is reached. When the algorithm reaches the maximum number of iterations, the algorithm ends and outputs the final optimal solution. The values ​​of each objective function are as follows: Figure 6A 、 6B and 6C.

[0218] Example 4

[0219] This embodiment is a system embodiment corresponding to the above method embodiment.

[0220] The cluster strategy-based joint planning system for container port berth and yard allocation includes:

[0221] The information collection module is used to collect information about ships, container trucks, and storage yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board the ship, the storage status of the storage yard, and the planned arrival time of the container trucks;

[0222] The optimal planning strategy solving module is used to input the collected information into a pre-built multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the yard container area, loading and unloading route, and loading and unloading time allocated to the container batch on the corresponding ship or container truck. The pre-built multi-objective mixed integer programming model includes three objective functions: minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and yard container areas, and minimizing the workload imbalance in the yard container area.

[0223] Actuators for executing vessel and yard operations according to optimal planning strategies.

[0224] Among them, the optimal planning strategy solving module selects the optimal solution based on the fuzzy correlation entropy fitness evaluation mechanism FCE-FEM when solving the multi-objective mixed integer programming model.

[0225] Each module or mechanism is mainly used to implement each step of the above method embodiment, which will not be described in detail here.

[0226] Example 5

[0227] This application also provides a computer-readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, a server, an App store, etc., storing a computer program that, when executed by a processor, implements corresponding functions. The computer-readable storage medium of this embodiment, when executed by a processor, implements the method for joint planning of container port berth and yard allocation according to the method embodiment.

[0228] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, or two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0229] The size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0230] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A method for joint planning of container port berth and yard allocation based on cluster strategy, characterized in that: The following steps are involved: S1. Collect information on ships, container trucks, and container yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board, the storage status of the container yard, and the planned arrival time of the container trucks. S2. Input the collected information into a pre-established multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the assigned container yard area, loading and unloading routes, and loading and unloading time for the corresponding container batch on the ship or truck. The pre-established multi-objective mixed integer programming model includes three objective functions, namely, minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and container yard areas, and minimizing the workload imbalance in the container yard areas. S3. Execute ship and yard operations according to the optimal planning strategy.

2. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to claim 1 is characterized in that: In step S1, the containers on the same ship or container truck are divided into multiple batches according to their destinations, and then the batches with the same destination are grouped into a container cluster. The container cluster is assigned to a container area template using an integer coding scheme, and then the container area template is assigned to a specific container area in the yard; the ship or container truck itself is used as the destination.

3. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to claim 1 is characterized in that: When solving multi-objective mixed integer programming models, the fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy is used to screen the optimal solution.

4. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to claim 3 is characterized in that: In the process of screening the optimal solution using the fuzzy correlation entropy-based fitness evaluation mechanism FCE-FEM, the global search capability is enhanced through an improved tabu search strategy. Specifically, based on the tabu search and multi-neighborhood search structure, a variety of neighborhood movement rules are designed to increase the neighborhood solution set, and the mutated neighborhood solution set is re-optimized.

5. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to claim 4 is characterized in that: Various neighborhood movement rules include the Exchange rule, the Fragment Insert rule, and the Intermingle Swap rule. The Exchange rule randomly selects two container slots and exchanges the genes of the two container areas. The Fragment Insert rule first randomly selects two container slots, removes the genes between the two container slot numbers, and then randomly selects a position in the remaining chromosomes to insert the removed genes into the position. The Intermingle swap rule randomly generates several container area positions and disrupts the gene order at the corresponding ship positions.

6. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to any one of claims 1 to 5, characterized in that: The constraints of the multi-objective mixed integer programming model include: each ship can only choose one berth to berth, two ships berthed at the same berth must have a berthing order, each ship can only berth and leave the berth once, the start and end times of loading and unloading of the ship are specified, the ship is required to start operations immediately after berthing and leave the berth immediately after finishing operations, and the loading and unloading operation time of the container batch is within the loading and unloading time of the corresponding ship.

7. The method for joint planning of container port berth and storage yard allocation based on cluster strategy according to any one of claims 1 to 5, characterized in that: When solving the mixed integer programming model, the objective function values ​​corresponding to each planning strategy are calculated in a specific order based on the three constructed objective function formulas. First, the ship's berthing time is added to the known port time to accurately calculate the ship's departure time, and then the cost incurred by the ship's deviation from the expected turnaround time interval is calculated according to the relevant formula; second, within the ship's berthing time range, the loading and unloading operation plan of the corresponding container batch is generated, and the workload imbalance within the planning range is calculated; finally, based on the loading and unloading operation plan and the ship's berthing position information, the storage location of each container batch in the yard is determined, and the average transportation distance of the container batch is calculated. Then, the average movement distance of all container batches is accumulated according to the formula, and the total container movement cost is finally obtained.

8. A container port berth and yard allocation joint planning operation system based on cluster strategy, characterized in that: include: The information collection module is used to collect information about ships, container trucks, and storage yards during the target time period, including the planned arrival time of the ship, the number and type of containers on board the ship, the storage status of the storage yard, and the planned arrival time of the container trucks; The optimal planning strategy solving module is used to input the collected information into a pre-built multi-objective mixed integer programming model to calculate the optimal planning strategy, including the berthing plan for all ships in a specified time period and the loading and unloading operation plan for each container batch. The berthing plan includes the ship's arrival time, berthing location, and berthing time. The loading and unloading operation plan for each container batch includes the yard container area, loading and unloading route, and loading and unloading time allocated to the container batch on the corresponding ship or container truck. The pre-built multi-objective mixed integer programming model includes three objective functions: minimizing the expected turnaround time for ship berthing, minimizing the total container transportation distance between berths and yard container areas, and minimizing the workload imbalance in the yard container area. Actuators for executing vessel and yard operations according to optimal planning strategies.

9. The container port berth and yard allocation joint planning operation system based on cluster strategy according to claim 8 is characterized in that: When solving a multi-objective mixed integer programming model, the optimal planning strategy solving module selects the optimal solution based on the fitness evaluation mechanism FCE-FEM based on fuzzy correlation entropy.

10. A computer storage medium, characterized in that A computer program executable by a processor is stored therein, and the computer program executes the container port berth and yard allocation joint planning operation method based on cluster strategy according to any one of claims 1 to 7.

Images