Container storage planning device, container storage planning system, and container storage planning method

Abstract

The present invention provides a container stacking planning apparatus for formulating a container stacking plan in a container terminal, comprising: a shipment sequence prediction unit that takes as input information representing attributes of containers previously unloaded at the container terminal and shipped out, i.e., container cargo information, and the actual shipment date of each of the previously shipped containers, a prediction model for predicting the shipment date of each container; and takes as input the container cargo information attached to containers unloaded at the container terminal and containers stacked at the container terminal, the prediction model to predict the shipment date of each container and assign a shipment sequence to each container; and The container storage planning department takes as input the container cargo information attached to the containers unloaded at the container terminal, the information indicating the storage location of containers in the container terminal (i.e., the storage container information), and the shipping sequence of each container calculated by the shipping sequence prediction department. It sets prescribed container storage restrictions to promote the efficiency of loading and unloading operations. Based on the shipping sequence of each container and the restrictions, it determines the storage location of each container in the container terminal in a manner that satisfies the evaluation indicators prescribed to minimize the transfer operations caused by shipping each container.

Description

Technical Field

[0001] This invention relates to a container stacking planning device, a container stacking planning system, and a container stacking planning method. Background Technology

[0002] With the development of the global economy, there is a huge demand for maritime transport that can transport large quantities of goods at low cost. Among them, maritime container transport has become a powerful means of achieving efficient cargo transportation by utilizing large-capacity containers such as 40-foot containers, which can be further transported by large trailers at the container terminal of the port of arrival and directly delivered to logistics warehouses.

[0003] At container terminals, a series of loading and unloading operations are carried out, following the process of unloading containers from berthing container ships, temporarily storing them in the container yard, and then transporting the stored containers out. Various solutions have been proposed to improve the efficiency of this loading and unloading operation, aiming to reduce the time containers spend at container terminals.

[0004] For example, Patent Document 1 discloses a container terminal system with the objective of "providing a distributed configuration system for loading and unloading machinery that effectively utilizes artificial intelligence, wherein the artificial intelligence is capable of pre-configuring efficient loading and unloading machinery for daily operations and outputting stacking instructions that reduce ineffective transshipment operations" (paragraph 0011). Furthermore, to achieve this objective, a "container terminal management system" is provided, comprising: a gantry crane for loading and unloading containers between the ship and the container terminal; yard trucks for transporting containers between the gantry crane and storage areas within the container yard; external trucks for transporting containers between the shipper and the storage areas; and a yard crane for loading and unloading containers between the storage areas and the yard trucks or the external trucks. This system possesses artificial intelligence constructed with a neural network, to which at least container-related information, container outbound information, and loading and unloading operation-related information are input. Using information and external factors of the terminal as input data, deep learning methods are used to determine the number and configuration status of yard cranes that maximize the utilization rate of the gantry cranes and minimize the waiting time for loading and unloading of external trucks. This information is then sent as output data, along with designated storage locations for each external truck and the yard trucks, and configuration instructions for each yard crane deployed according to the minimum required number (claim 1). In particular, "the output data may also include a storage scheme that minimizes the predicted number of container transfers until the stored containers are shipped out" (claim 2).

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent Application Publication No. 2019-104577 Summary of the Invention

[0008] The technical problem that the invention aims to solve

[0009] In container yards, due to space limitations, containers are typically stacked in multiple layers for storage. If other containers are stacked on top of the container that should be shipped out first, these other containers need to be transferred to another location before the priority container is shipped out, resulting in a so-called "repositioning" operation. Patent Document 1 proposes a technical solution to achieve efficient loading and unloading operations by generating a container stacking scheme that minimizes the number of repositioning operations.

[0010] However, in order to maximize the efficiency of actual container terminal loading and unloading operations, simply planning container stacking in a way that theoretically minimizes the number of transfers may not necessarily result in an efficient stacking plan due to factors such as on-site operational constraints. For example, even if a container stacking plan that minimizes the number of transfers is generated, if containers are stacked in a manner that is difficult to operate with cranes when using transfer cranes to move the stacked containers out, or if containers of different capacities are stacked in the same bay, the outbound operation will be inefficient. In addition, the priority of container outbound operations may sometimes change due to external factors such as the late arrival of outbound trailers, resulting in unexpected transfers and still causing inefficiencies in outbound operations.

[0011] The present invention was made in view of the following background, and its object is to provide a container stacking planning device, a container stacking planning system and a container stacking planning method that can generate efficient container stacking plans that are conducive to loading and unloading operations, taking into account various conditions required at the container terminal site.

[0012] Technical solutions for solving the problem

[0013] One aspect of the present invention for solving the above-mentioned problems is a container stacking planning apparatus for formulating a container stacking plan in a container terminal, comprising: a shipment sequence prediction unit that takes as input information representing attributes of containers previously unloaded to the container terminal and shipped out, i.e., container cargo information, and the actual shipment date of each of the previously shipped containers, a prediction model for predicting the shipment date of each container, and takes as input the container cargo information attached to containers unloaded to the container terminal and containers stacked at the container terminal the prediction model to predict the shipment date of each container and assign each container a... The shipping sequence prediction unit and the container stacking planning unit take as input the container cargo information attached to the containers unloaded to the container terminal, the information indicating the stacking location of containers in the container terminal, i.e., the stacked container information, and the shipping sequence prediction unit's calculation of the shipping sequence order of each container. They set prescribed container stacking restrictions to promote the efficiency of loading and unloading operations. Based on the shipping sequence of each container and the restrictions, they determine the stacking location of each container in the container terminal in a manner that satisfies the evaluation indicators prescribed to minimize the transfer operations caused by shipping out each container.

[0014] Another aspect of the present invention is a container stacking planning system for formulating and utilizing container stacking plans in a container terminal, comprising a container stacking planning device and a terminal device, wherein the container stacking planning device comprises: a container terminal business system storing: information representing attributes of containers previously unloaded to the container terminal and shipped out, i.e., container cargo information; information including the unloading sequence of unloaded containers, i.e., unloading container information; and information representing the stacking locations of containers stacked at the container terminal, i.e., stacked container information; and a shipment sequence prediction unit that takes the container cargo information of containers previously unloaded to the container terminal and shipped out and the actual shipment date of each previously shipped container as input, generates a prediction model for predicting the shipment date of each container, and loads unloaded to the container terminal... The system uses the container cargo information attached to the containers and containers stored at the container terminal as input to the prediction model to predict the shipping date of each container and assign a shipping order to each container; and the container storage planning unit takes the unloading container information, the stored container information, and the shipping order of each container calculated by the shipping order prediction unit as input, sets prescribed container storage restrictions to promote efficient loading and unloading operations, and determines the storage location of each container in the container terminal based on the shipping order of each container and the restrictions, in a manner that satisfies the evaluation indicators prescribed to minimize the transfer operations caused by shipping each container. The terminal device receives the container storage plan formulated by the container storage planning device and displays the container storage plan.

[0015] Another aspect of the present invention is a container stacking planning method for formulating a container stacking plan in a container terminal, wherein an information processing unit having a computing device and a storage device performs the following steps: taking as input information representing attributes attached to containers that have been unloaded to the container terminal and shipped out in the past, i.e., container cargo information, and the actual shipping date of each of the containers shipped out in the past, a prediction model for predicting the shipping date of each container; and taking as input the container cargo information attached to the containers unloaded to the container terminal and the containers stacked at the container terminal, the prediction model is used to predict the shipping date of each container. The process includes assigning a shipping order to each container; and taking as input the container cargo information attached to the container unloaded at the container terminal, information indicating the container storage location in the container terminal (i.e., storage container information), and the shipping order of each container calculated by the shipping order prediction unit, setting prescribed container storage restrictions to promote efficient loading and unloading operations, and determining the storage location of each container in the container terminal based on the shipping order of each container and the restrictions, in a manner that satisfies evaluation indicators prescribed to minimize the transfer operations generated by shipping each container.

[0016] The effects of the invention

[0017] According to the present invention, it is possible to generate an efficient container stacking plan that is conducive to loading and unloading operations, taking into account various conditions required by container terminals.

[0018] Other problems, structures, and effects not described above will become clear from the following description of the implementation methods. Attached Figure Description

[0019] Figure 1 This is a schematic top view illustrating an example of the equipment and functions of a container terminal.

[0020] Figure 2 This is a schematic diagram illustrating containers stacked in shells.

[0021] Figure 3 This is a schematic diagram illustrating the unloading and stacking of containers.

[0022] Figure 4A This is a schematic diagram illustrating the container stacking status of a container yard.

[0023] Figure 4B This is a schematic diagram illustrating the container stacking status of a container yard.

[0024] Figure 5 This is an illustration of the number of unfavorable events used in container stacking planning.

[0025] Figure 6 This is a block diagram illustrating a structural example of a container stacking planning system that includes a container stacking planning device according to an embodiment of the present invention.

[0026] Figure 7 This is a block diagram illustrating an example of the hardware structure of the container stacking planning device according to this embodiment.

[0027] Figure 8 This is a block diagram representing the structure of a container terminal business system 10.

[0028] Figure 9 This is a block diagram illustrating a structural example of a container stacking planning device 20.

[0029] Figure 10 This is a diagram illustrating a structural example of container cargo information 14.

[0030] Figure 11 This is a diagram illustrating the structure of unloading container information 15.

[0031] Figure 12 This is a diagram illustrating a structural example of stacked container information 16.

[0032] Figure 13 This is a diagram illustrating a structural example of container stacking plan information 25.

[0033] Figure 14 This is a diagram representing a structural example of constraint 23.

[0034] Figure 15 This is a diagram illustrating the structure of evaluation index 24.

[0035] Figure 16A This is a flowchart illustrating an example of the data processing flow for the outgoing prediction model learning process in this embodiment.

[0036] Figure 16B This is a flowchart illustrating an example of the data processing flow for the outgoing prediction model learning process in this embodiment.

[0037] Figure 17A This is a flowchart illustrating an example of the data processing flow for the container stacking plan formulation process in this embodiment.

[0038] Figure 17B This is a flowchart illustrating an example of the data processing flow for the container stacking plan formulation process in this embodiment.

[0039] Figure 18 This is a flowchart illustrating an example of the data processing flow for container stacking and selection processing in this embodiment.

[0040] Figure 19This is a flowchart illustrating an example of the data processing flow for selecting container storage locations in this embodiment. Detailed Implementation

[0041] Hereinafter, with reference to the accompanying drawings, embodiments of the container stacking planning apparatus, container stacking planning system, and container stacking planning method of the present invention will be described.

[0042] Before proceeding, we will first describe the equipment and functions of a typical container terminal, which is the premise of this explanation.

[0043] <Overview of Container Terminals>

[0044] Figure 1 This is a schematic top-view diagram illustrating the equipment and functions of a typical container terminal. A container terminal is constructed as part of the harbor where container ships (CS) can enter, allowing them to berth at the quay wall (PI). At the quay wall (PI), a gantry crane (GC) is installed, movable along the quay wall, to unload containers from the CS. On the land side of the gantry crane (GC), a container yard (CY) is located, serving as a temporary storage area for unloaded containers (CNTs) until they are shipped from the container terminal. Within the container yard (CY), roughly rectangular areas called blocks (BLs) are provided, where large numbers of containers (CNTs) are stacked as groups. The columns formed by arranging blocks (BLs) along their length are called lanes (LNs). Between lanes (LNs), transport lanes (CLNs) are provided, serving as passageways for in-yard transport vehicles (YDCHs) that transport containers (CNTs) within the container terminal. In each lane (LN), a transfer crane (TRC) is installed, capable of moving along the length of a block (BL). Each transfer crane (TRC) stacks and retrieves container CNTs from and from the block BL belonging to its assigned lane (LN). Each block BL is divided along its length into multiple areas called bays (bays). In this embodiment, the decision of which bay in which block BL to store unloaded container CNTs, and how to store them in each bay, is evaluated under various constraints. The gate (GT) is a facility used to manage the outbound movement of container CNTs from the container terminal, which is also a bonded area.

[0045] The process from unloading a container (CNT) to its exit from the gate is roughly as follows. Additionally, the numbers in parentheses below... Figure 1 The numbers in parentheses correspond to the given numbers.

[0046] (1) First, the gantry crane GC takes out the containers from the container ship CS in the prescribed order and stacks them on the yard transport vehicle YDCH for transport within the container terminal.

[0047] (2) Then, the YDCH is transported by the on-site transport vehicle to the block BL of the pre-designated lane LN.

[0048] (3) In the designated block BL, the transfer crane TRC stacks the container CNT to the designated location of the pre-designated BLBY.

[0049] (4) Containers stored in Bay BY are taken out by the transfer crane TRC according to the priority of shipment and stacked on land transport trailers, etc., and transported out of the gate GT.

[0050] Traditionally, for each container (CNT), the decision on which warehouse and location to transport and stack it is made manually, taking into account various factors such as shipping priority and based on years of experience.

[0051] <Storing Containers by>

[0052] exist Figure 2 This schematic diagram shows a front view (cross-section along the width of block BL) of a container bay (CNT) for stacking containerized container towers (CNTs). Within a bay, CNTs are stacked in a matrix configuration. Figure 2 In this example, the bay BY has 6 rows in the width direction and 5 tiers in the height direction, allowing a maximum of 6 × 5 = 30 container NTs to be stacked in one bay BY. Figure 2 In the diagram, shaded areas represent stacked containers, while dashed lines represent areas where containers have not yet been stacked.

[0053] In columns 1 through 5, two layers of containers are stacked up to the second layer; only in column 6 are five layers of containers stacked up to the fifth layer. Therefore, for example, the unloading area for the next container can be located in column 3, layer 3, indicated by the arrow. In this way, the container stacking area, when a specific section is specified, can be represented by a coordinate system consisting of a combination of columns and layers. For example... Figure 2 The next unloading location is denoted as (3, 3).

[0054] <The Idea of ​​a Container Storage Plan>

[0055] Here, the basic idea of ​​the container stacking plan of the present invention will be explained. Figure 3 This schematically illustrates the unloading of containers and their storage in the container yard.

[0056] Figure 3This diagram illustrates three containers stacked on a container ship awaiting unloading and five containers already unloaded and stored in a specific container yard. The numbers labeled on each container indicate their priority for shipment after stacking. Container number 1 has the highest priority, with containers numbered 2, 3, and so on, decreasing in priority.

[0057] exist Figure 4A , Figure 4B In the middle, it is indicated that in Figure 3 In this state, three containers are unloaded from the container ship and stacked into the corresponding bays. First, refer to... Figure 4A If, without considering the shipping priority of each container, they are stacked two layers high in each column, in other words, the number of layers is 2. It is then known that, when retrieving containers according to the shipping priority, a transfer operation is required: moving container 5 first to retrieve container 1, and moving container 7 first to retrieve container 4. Clearly, these transfer operations reduce the efficiency of loading and unloading operations and are not optimal.

[0058] On the other hand, refer to Figure 4B Because containers are stacked in order of shipment priority, containers 1 through 5 can be retrieved without relocation; only the retrieval of container 6 requires relocation of container 7. However, compared to... Figure 4A Compared to the previous situation, loading and unloading operations are obviously more efficient.

[0059] Based on a container stacking plan that takes into account the shipping priorities described above, this implementation method incorporates the evaluation metric of "number of adverse events." Figure 5 The concept of penalty count is explained below. Penalty count can be defined as "the number of containers stacked on top of a container with a higher shipping priority but lower shipping priority compared to that container." When this concept of penalty count is applied... Figure 4A , Figure 4B In the example of container stacking, they respectively become Figure 5 Cases like those in Examples 1 and 2. That is, in... Figure 4A In the example of container stacking, because two containers with lower shipping priority are stacked on top of a container with higher shipping priority, the number of unfavorable events is 2. Conversely, in... Figure 4BIn the example of container stacking, since one container with a lower dispatch priority is stacked on top of a container with a higher dispatch priority, the number of adverse events is 1. In this embodiment, the container stacking plan is formulated in a way that minimizes the number of adverse events for a given container. Furthermore, regarding the selection of containers, the container stacking plan is formulated in a way that minimizes the number of transfers for the entire container terminal, under conditions such as prioritizing containers with fewer stacked containers. In addition, indicators other than the number of adverse events can be used while minimizing transfer operations. For example, the container stacking plan can be formulated in a way that minimizes the total time required for the resulting transfer operations while taking into account dispatch priority.

[0060] <Container Stacking Planning System>

[0061] Next, the structure and function of the container stacking planning system of this embodiment will be explained. Figure 6 This illustrates the overall structure of the container stacking planning system 1 in this embodiment.

[0062] like Figure 6 As shown, the container stacking planning system 1 of this embodiment includes a container terminal business system 10, a container stacking planning device 20, terminal devices 30A and 30B, and a communication network 40 that communicatively connects the constituent elements.

[0063] The container terminal operation system 10 has the function of comprehensively managing information about containers arriving at the container terminal, providing basic information about each container as a prerequisite for formulating a container stacking plan. The container stacking plan device 20 has the function of formulating the container stacking configuration for each bay of the container yard based on the information about each unloaded container provided by the container terminal operation system 10. As described later, the container stacking plan device 20 can formulate a container stacking plan that considers outbound priority and relies on more efficient loading and unloading operations based on an outbound prediction model with learning capabilities such as deep learning.

[0064] Terminal devices 30A and 30B include display devices that output the storage coordinates of containers according to the container storage plan formulated by the container storage planning device 20. These coordinates consist of the container bays where each container should be stored and the storage locations within those bays, i.e., rows and layers. For example, these devices may be installed on the cabs of loading and unloading equipment such as yard transport vehicles or transfer cranes. The communication network 40 is a wired or wireless communication line that allows communicative connections between the container terminal business system 10, the container storage planning device 20, and the terminal devices 30A and 30B, such as the Internet, dedicated lines, WAN (Wide Area Network), or LAN (Local Area Network).

[0065] <Hardware Structure of Container Stacking Planning System>

[0066] Next, the hardware structure of the container terminal business system 10, container storage planning device 20, and terminal devices 30A and 30B that constitute the container storage planning system 1 will be described. Figure 7 This describes an example of the hardware structure of an information processing device, which is employed as one of these components and has communication capabilities via the communication network 40. The container terminal business system 10 and the container stacking planning device 20 can be, for example, equipped with… Figure 7 The information processing device 100 illustrated herein is configured as a server computer, and the terminal devices 30A and 30B can also function as a server computer. Figure 7 The structure shown in the example consists of a personal computer or tablet computer terminal device.

[0067] The information processing device 100 includes a computing unit 110, a main storage unit 120, an auxiliary storage unit 130, an input device 140, an output device 150, and a communication device 160. The computing unit 110 can be composed of a processor such as an MPU or CPU. The main storage unit 120 provides storage areas used by the computing unit 110 and includes storage devices such as ROM, RAM, and flash memory. The auxiliary storage unit 130 provides storage areas for programs executed by the computing unit 110 and data used by these programs, including storage devices such as hard disk drives (HDDs), semiconductor drives (SSDs), optical drives, and USB storage devices. The input device 140 includes input devices capable of inputting data to the container stacking planning system 1 via the information processing device 100, such as a keyboard, mouse, touch panel, and voice input device. The output device 150 is an output device capable of outputting data from the container stacking planning system 1 and can include a monitor, printer, and voice output device. The communication device 160 can be composed of a communication module such as a network interface card. The various elements within the information processing device 100 can be communicatively connected via internal data communication lines (not shown in the diagram).

[0068] In addition, the container terminal business system 10 and the container stacking planning device 20 may not be constructed as separate components like the information processing device 100, but may be implemented as the same component. For example, in a virtual computer or cloud system environment, a parallel distributed processing device, a blockchain or other distributed storage, or a ledger system may be constructed as a reference source.

[0069] Next, the functions of the container terminal business system 10 and the container stacking planning device 20 will be explained.

[0070] <Container Terminal Business System 10>

[0071] Figure 8A block diagram illustrates the functions of the container terminal business system 10 in this embodiment. As described with respect to the container storage planning system 1, the container terminal business system 10 is a system that comprehensively manages various information regarding containers that have previously arrived at the container terminal for unloading and shipment, and containers scheduled to arrive at the container terminal in the future. Figure 8 As illustrated, the container terminal business system 10 has a data extraction unit 11, a data processing unit 12, and a data I / O interface unit 13, as well as container cargo information 14, unloading container information 15, and stacked container information 16, which are stored in the data storage area.

[0072] The data extraction unit 11, for example, has the function of extracting the required data from the container cargo information 14, the unloading container information 15, and the stacked container information 16 based on a request from the container stacking planning device 20 (described later). The data extraction unit 11 can, for example, be configured as a database management system (DBMS) storing the database of the container cargo information 14, the unloading container information 15, and the stacked container information 16. Furthermore, an example of the structure of the container cargo information 14, the unloading container information 15, and the stacked container information 16 will be described later.

[0073] The data processing unit 12 has the function of processing the data extracted by the data extraction unit 11 into a data format suitable for input to the container stacking planning device 20, which will be described later. This data processing includes checking and correcting missing, invalid, and illegal data in the extracted learning data.

[0074] The data I / O interface unit 13 has an interface function that sends the data processed by the data processing unit 12 to the container stacking planning device 20 via the communication network 40.

[0075] The data extraction unit 11, the data processing unit 12, and the data I / O interface unit 13 are composed of Figure 7 The program installed by the information processing device 100 in the example can also be partially or entirely implemented using hardware. Furthermore, the container terminal business system 10 may also include function blocks for implementing other functions besides those described above.

[0076] <Container Stacking Planning Device 20>

[0077] Figure 9The function of the container stacking planning device 20 in this embodiment is illustrated using a block diagram. As described with respect to the container stacking planning system 1, the container stacking planning device 20 has the function of determining the container stacking pattern of each container in the container yard based on information provided by the container terminal business system 10 regarding each container that has been unloaded in the past or is to be unloaded now.

[0078] like Figure 9 As illustrated, the container stacking planning device 20 has an outbound prediction model 21, a container stacking planning model 22, and a data I / O interface unit 26, as well as container stacking planning information 25, which is stored as data in the data storage area.

[0079] The shipment prediction model 21 has the following functions: It obtains the actual shipment dates of containers unloaded at the container terminal in the past from the container terminal business system 10, uses deep learning and other technologies to learn from the information belonging to each container, and predicts the shipment date of a container based on the information about containers to be unloaded at the container terminal in the future, using the learned model. In this embodiment, the shipment prediction model 21 is also described as a model obtained using machine learning or deep learning; however, the method of formulating the prediction model is not limited to this. For example, a prediction model derived by analyzing the parameters of the actual shipment dates using statistical methods such as correlation analysis can also be used.

[0080] The container stacking plan model 22 has the function of determining the shipping priority of containers according to the predicted shipping dates of each container obtained from the shipping forecast model 21, and formulating a container stacking plan for the containers that can be shipped out of the container according to the shipping priority. In the container stacking plan model 22, the constraints 23 that should be considered when formulating the stacking plan and the evaluation indicators 24 used to evaluate the formulated container stacking plan are stored as data.

[0081] The container stacking plan information 25 is stored as the container stacking plan formulated by the container stacking plan model 22. Based on the request of the container stacking plan model 22, the container stacking plan information 25 is sent to the container terminal business system 10 and terminal devices 30A and 30B via a data I / O interface unit 26 that has an interface function with the communication network 40.

[0082] Outbound prediction model 21, container stacking planning model 22, and data I / O interface unit 26 are used as... Figure 7 The information processing device 100 in this example executes a program installation, although some or all of it can also be implemented using hardware. Furthermore, the container stacking planning device 20 may also include function blocks for implementing other functions besides those described above.

[0083] Next, the information on container cargo 14, unloading container 15, stacked container 16, and container stacking plan 25 will be explained.

[0084] <Container Cargo Information 14>

[0085] Figure 10 This is a structural example representing container cargo information 14. Figure 10 In the example, container cargo information 14 is presented in tabular form, but it can also be stored in other data formats. The same applies to unloading container information 15, stacking container information 16, and container stacking plan information 25, which will be discussed later.

[0086] Container cargo information 14 records the following items: container number (used as a unique identifier for each container), container arrival date, arrival time, container departure date, departure time, consignee, cargo name (indicating the contents of the container), carrier (indicating the merchant shipping the container from the container terminal), ocean freight forwarder (indicating the merchant responsible for the maritime transport of the container), destination name (indicating the destination of the departing container, such as a logistics warehouse), container dimensions, container type (indicating the purpose of the container, such as dry cargo or refrigerated containers), and vessel code (used to uniquely identify the container ship carrying the container). Figure 10 The contents of the items included in the container cargo information 14 are illustrated, but they are merely illustrative for ease of understanding of the invention and do not limit this embodiment. The same applies to the unloading container information 15, the stacking container information 16, and the container stacking plan information 25, which will be described later.

[0087] In this embodiment, as described later, a shipping prediction model for predicting the shipping date of each container is generated by using the data of the container cargo information 14 as learning data. The shipping date prediction process for that container is then performed by inputting the container cargo information 14 attached to the container being predicted into this prediction model. Furthermore, the items recorded in the container cargo information 14 can be used in their entirety or only in part during the prediction process. Additionally, the container cargo information 14 may also include... Figure 10 Items other than those listed in the examples.

[0088] <Unloading Container Information 15>

[0089] Figure 11Example of a structure representing unloading container information 15. Unloading container information 15 is data indicating the order in which containers are unloaded from a container ship arriving at the container terminal. Unloading container information 15 is shared, for example, among operators of gantry cranes unloading containers from container ships, yard transport vehicles carrying unloaded containers to the container yard, and transfer cranes that remove containers from yard transport vehicles and stack them to designated bays in the container yard.

[0090] The unloading container information 15 records the container number, the vessel code, and the unloading sequence, indicating the order of unloading of containers from the container ship specified by the corresponding vessel code.

[0091] <Storage Container Information 16>

[0092] Figure 12 This is a structural example representing stacked container information 16. Stacked container information 16 is generated using... Figure 1 The combination of aisles, blocks, bays, columns, and layers in the example container yard represents data on the storage locations of each container in each bay of the container yard. Figure 12 The stacked container information 16 records the unique container number that identifies each stacked container in conjunction with the aforementioned aisles, blocks, bays, columns, and layers. The content of the stacked container information 16 is updated approximately in real time in response to container unloading, shipping operations, and the resulting transfer operations.

[0093] <Container Storage Plan Information 25>

[0094] Figure 13 This is a structural example of container storage plan information 25. Container storage plan information 25 records the locations within the container yard where containers unloaded from container ships should be stored, along with their dispatch priority. Specifically, in... Figure 13 In the example, the container stacking plan information 25 records the container number, the aisle, block, bay, column, and layer combination indicating the location where the container should be stacked, and the container's shipping priority. The container stacking plan information 25 serves to inform loading and unloading operators of the unloading sequence, for example, by displaying outputs on the dashboards of loading and unloading equipment such as yard transport vehicles and transfer cranes.

[0095] Next, the constraints and evaluation indicators used by the container stacking planning model 22 of the container stacking planning device 20 will be explained.

[0096] <Restriction 23>

[0097] Figure 14This represents an example of setting constraint 23. Constraint 23, when used in the container storage planning model 22 to formulate a container storage plan according to shipping priority, specifies various conditions to be considered during loading and unloading operations, such as constraints related to safety and improving operational efficiency. Constraint 23 extracts items that should be pre-defined and sets them in the container storage planning model 22.

[0098] The following illustrates the limitation condition 23 set in this embodiment. Figure 14 ).

[0099] • Prioritize the use of columns with fewer containers.

[0100] • Prioritize the use of containers with fewer storage units.

[0101] • The number of containers that can be stacked inside the oyster is limited based on the maximum stackable number of containers in that oyster (e.g., 80% of the maximum stackable number).

[0102] These restrictions aim to prevent containers from concentrating in specific bays or columns.

[0103] • Containers should not be stacked on the top layer of the storage area.

[0104] • The storage of containers within the container yard does not create "valley".

[0105] • The stacking of containers inside the container house results in a relatively low number of trucks on the side of the transport lane.

[0106] • No independent container stacking towers without adjacent containers will be built in Bene.

[0107] • In cases where there are no containers stored in the container yard, priority should be given to using the column on the opposite side of the transport lane.

[0108] These restrictions are set to prevent a decrease in the operational efficiency of loading and unloading equipment such as transfer cranes used for container stacking and retrieval. For example, the condition that container stacking inside the yard should not create "valleys" is set to prevent transfer cranes from having difficulty handling containers in the inner rows when containers are stacked high at both ends of the rows inside the yard, thus creating "valleys." Furthermore, the condition that priority should be given to using rows on the opposite side of the transport vehicle aisle when there are no containers stacked inside the yard is to clear the transport vehicle aisle side for transport vehicles, thereby improving the operational efficiency of transfer cranes.

[0109] • Stack containers of the same size in the same bay.

[0110] • Do not store containers of different categories in the same bay.

[0111] • Do not store incoming containers and outgoing containers in the same container stack.

[0112] These constraints achieve the effect of maximizing the number of containers stored in a treasury while homogenizing loading and unloading operations by ensuring that the containers stored in the same treasury are of the same shape / type (dry, frozen, etc.) / purpose.

[0113] For each item in restriction condition 23, weighting can also be set according to which condition is emphasized when formulating the container storage plan.

[0114] The restrictions that should be set are not subject to Figure 14 In addition to the aforementioned limitations, appropriate provisions can be made based on the specific circumstances of each container terminal.

[0115] <Evaluation Indicator 24>

[0116] Figure 15 This illustrates an example of setting evaluation index 24. Evaluation index 24 is an index used when evaluating a container storage plan developed using container storage plan model 22. Figure 15 In this example, two evaluation indicators are set as evaluation metrics in this implementation: "priority is given to storage plans with fewer adverse events" and "priority is given to plans farther from the transport vehicle aisle side." Regarding the number of adverse events, for example... Figure 5 As explained, after loading the same container, when retrieving containers according to the shipping priority order, the fewer the number of containers stacked on top of containers with a higher shipping priority order compared to those with a lower shipping priority order, i.e., the fewer unfavorable events, the higher the evaluation of the stacking plan. The evaluation criterion of "priority for containers farther from the transport vehicle aisle" is explained in relation to constraint 23.

[0117] Similar to constraint 23, weights can also be assigned to each item in evaluation index 24 based on which index is prioritized when developing the container storage plan.

[0118] The evaluation indicators that should be set are not affected Figure 15 In addition to the aforementioned limitations, appropriate provisions can be made based on the specific circumstances of each container terminal.

[0119] Next, the data processing performed by the container stacking planning system 1 with the structure described above will be explained.

[0120] <Data Processing of Container Stacking Planning System 1>

[0121] First, the outbound prediction model learning process, which is jointly executed by the container terminal business system 10 and the container stacking planning device 20, is explained.

[0122] [Output Prediction Model Learning Processing]

[0123] exist Figure 16A , Figure 16B The following describes the data processing flow for the outbound prediction model learning process of the container stacking planning system 1 of this embodiment. This outbound prediction model learning process is performed to construct the outbound prediction model 21 of the container stacking planning device 20 before running the container stacking planning system 1 of this embodiment. Furthermore, in the following description, the term "outbound date" refers to the date and time of outbound shipment.

[0124] Reference Figure 16A First, in the container terminal business system 10, the data extraction unit 11 extracts the container cargo information recorded in the container cargo information 14, which includes information about containers that have been unloaded in the past and information about containers stored in the container yard. After the data processing unit 12 processes the data, it is sent from the data IO interface unit 13 to the container stacking planning device 20 (S10, S11).

[0125] Next, in the container terminal business system 10, the data extraction unit 11 extracts the actual shipping date from the container cargo information recorded in the container cargo information 14 regarding the container that was unloaded in the past. After the data processing unit 12 processes the data, it is sent from the data IO interface unit 13 to the container stacking planning device 20 (S13, S14).

[0126] In the container stacking planning device 20, the data I / O interface unit 26 sends the container cargo information about containers that have been unloaded in the past and the container cargo information about containers stacked in the container yard to the outbound prediction model 21 (S12). Data about the actual outbound date of containers that have been unloaded in the past is also sent to the outbound prediction model 21 (S15).

[0127] Next, refer to Figure 16B The container stacking planning device 20 selects an algorithm from a plurality of pre-installed algorithms for learning functions to predict the shipping priority (shipment date) (S16). As the algorithm to be adopted, machine learning algorithms such as RNN (Recurrent Neural Network) and deep learning are considered.

[0128] Next, the shipping prediction model 21, based on the selected algorithm, uses pre-prepared analysis parameters for multiple modes, and performs shipping date prediction learning for each mode, based on container cargo information obtained from the container terminal business system 10 as the learning tool and the actual shipping date (S17). Additionally, the analysis parameters include, for example, hyperparameters such as the number of hidden layers, which are set to adjust the deep learning network.

[0129] The shipping prediction model 21 compares the predicted shipping date with the actual shipping date recorded in the past container cargo information based on past container cargo information (S18). Furthermore, the shipping prediction model 21 completes this series of processes by making the comparison result of the difference the smallest possible analytical parameter as its logical selection (S19).

[0130] Based on the shipment prediction model 21 constructed as described above, the shipment date of unloaded containers can be predicted with high accuracy using a prediction model obtained by learning or statistically analyzing past container cargo information. Therefore, the shipment priority order of each container can be obtained more accurately based on its prediction results.

[0131] [Container Storage Planning and Processing]

[0132] Next, the container storage planning process implemented using the outbound prediction model 21 and the container storage planning model 22 of the container storage planning device 20 will be explained. Figure 17A , Figure 17B The data processing flow for container stacking planning in this embodiment is illustrated below.

[0133] Reference Figure 17A First, the data extraction unit 11 of the container terminal business system 10 extracts cargo information about each unloading container from the container cargo information 14, unloading container information 15, and stacked container information 16. After the data processing unit 12 performs the required data processing, it sends the data from the data IO interface unit 13 to the data IO interface unit of the container stacking planning device 20 (S20, S21).

[0134] In addition, the data extraction unit 11 extracts information about each unloading container from the unloading container information 15 and the stacked container information 16. After the data processing unit 12 performs the necessary data processing, it sends the data from the data IO interface unit 13 to the data IO interface unit of the container stacking planning device 20 (S25, S26, S27).

[0135] Then, in the container stacking planning device 20, the data I / O interface unit 26 sends the received cargo information about unloaded containers and stacked containers to the shipping prediction model 21 (S22). Based on the received cargo information of the unloaded and stacked containers, the shipping prediction model 21 predicts the shipping date of each container (S23) and assigns a shipping priority to each container. The shipping priority can be, for example, based on... Figures 3-5 A series of numbers as shown (S24).

[0136] Next, refer to Figure 17BThe data I / O interface unit 26 of the container stacking planning device 20 sends the received unloading container information and stacking container information to the container stacking planning model 22 (S28). In addition, the container stacking planning model 22 obtains the constraints 23 and evaluation indicators 24 used in the stacking planning process (S29, S30).

[0137] Based on the acquired information on stacked containers, unloading containers, and the shipping priority, constraints, and evaluation indicators calculated by the shipping prediction model 21, the container stacking planning device 20 first determines the bay for each container (S31). Then, the container stacking planning model 22 further determines the stacking coordinates within the bay determined in S31 as a combination of columns and layers, based on the acquired information on stacked containers, unloading containers, and the shipping priority, constraints, and evaluation indicators calculated by the shipping prediction model 21 (S32).

[0138] The container stacking plan model 22 repeatedly determines the stacking coordinates of all containers in the container yard (S33, No). When it is determined that the number of unfavorable calculations for all containers has converged to a minimum (S33, Yes), the container stacking plan formulation process ends. The formulated container stacking plan data is stored as container stacking plan information 25 in the container stacking plan device 20 and sent to the container terminal business system 10 and terminal devices 30A and 30B via the data I / O interface unit 26. At the terminal devices 30A and 30B, for example, the output device such as a monitoring display... Figure 13 The container stacking plan information is displayed in the form of charts, as shown in the example, thereby enabling the unloading and stacking operation steps of containers to be communicated to the operators of loading and unloading equipment.

[0139] [Specific examples of container storage planning and processing]

[0140] Here, based on the above explanation, the data processing performed by the container stacking planning model 22 of the container stacking planning device 20 will be described. First, in Figure 18This example illustrates the data processing flow for container storage bay selection. When data processing begins in S40, the container storage planning model 22 determines the bay to be used for storing the container based on storage container information, unloading container information, shipping priority, and evaluation indicators (S41). The container storage planning model 22 then checks whether the determined bay violates the bay selection constraints (S42). If a violation is found, it returns to the bay selection step in S41. If the bay selection constraints are not violated (S42, No), the container storage planning model 22 determines whether bay selection is permissible according to the evaluation indicators (S43). If permissible (S43, Yes), the data processing ends (S44). If permissible according to the evaluation indicators (S43, No), the container storage planning model 22 returns to the bay selection step in S40.

[0141] Based on the above-described storage bay selection process using container storage planning model 22, it is possible to select storage bays that meet the evaluation criteria while considering the priority of shipment, under the pre-defined constraints of the storage plan.

[0142] Next, refer to Figure 19 The illustrated data processing flow for storage location selection illustrates the storage location selection process using container storage planning model 22.

[0143] When the container storage planning model 22 begins data processing in S50, it first determines the storage location for each container based on the information of stored containers, unloaded containers, and shipping priority (S51). Next, for all columns of that container, the container storage planning model 22 calculates the number of adverse events based on evaluation indicators and constraints (S52, S53). After completing the processing of all columns in S53, the container storage planning model 22 determines whether the number of adverse events for that container has converged to the minimum (S54). If convergence is determined (S54, Yes), the data processing ends. If it is determined in S54 that the number of adverse events has not converged to the minimum (S54, No), the container storage planning model 22 returns to the processing in S50 to determine the storage location again.

[0144] Based on the above storage location selection process, under specified constraints and evaluation indicators, the storage location of containers can be determined in a way that minimizes the number of adverse events in each container, preventing the efficiency loss caused by container transfer operations during shipment, and enabling containers to be stacked in a highly efficient manner suitable for unloading operations.

[0145] The above provides a detailed description of how to implement the present invention. However, the present invention is not limited thereto and various modifications can be made without departing from its spirit.

[0146] For example, the evaluation metrics mentioned above can include the scenario where the number of containers stacked on top of containers with higher shipping priority but lower shipping priority is minimized. This minimizes the repositioning operations required to move containers stacked on top of containers with higher shipping priority but lower shipping priority.

[0147] Furthermore, among the aforementioned constraints, the stacking arrangement of containers at the container terminal should not obstruct the stacking and retrieval of containers by the terminal's loading and unloading equipment. This constraint improves the efficiency of container stacking and retrieval operations, reducing loading and unloading time. Moreover, if the aforementioned constraints are included within the smallest unit area constituting the specified number of containers to be stacked at the container terminal, and containers with the same specified attributes are stacked in the same unit, the number of containers stacked in the same unit can be maximized, and container loading and unloading operations can be homogenized, preventing containers from exceeding the limits of the transport vehicle aisles and obstructing the passage of loading and unloading equipment.

[0148] Furthermore, the aforementioned container storage planning department can, according to the aforementioned constraints, select the bays within the smallest unit area of ​​the container terminal that constitutes the storage of the prescribed number of containers, based on the bays that best meet the evaluation criteria for each container to be stored. Thus, for example, it can promote efficient utilization of bays by prioritizing container stacking from bays that do not store containers or bays with a small number of containers stored.

[0149] Furthermore, the container stacking planning department, when selecting a stacking location that best meets the aforementioned evaluation criteria for each container to be stacked, can efficiently stack and retrieve containers for each container, provided that the selected container storage facility is stacked according to the aforementioned constraints. In this case, if the evaluation criteria include minimizing the number of containers stacked on top of containers with higher shipping priority compared to those with lower shipping priority, then the container handling operations can be minimized, maximizing the efficiency of container loading and unloading.

[0150] Furthermore, if the aforementioned shipment order prediction unit compares the predicted shipment date of each container, which is based on the past shipment information, with the actual shipment date recorded in the past shipment information, and generates the prediction model used in the shipment order prediction process in a way that minimizes the difference in the comparison, then the shipment priority used in the formulation of container stacking plans can be predicted with higher accuracy.

[0151] Explanation of reference numerals in the attached figures

[0152] 1 Container Stacking Planning System

[0153] 10 Container Terminal Business System

[0154] 11 Data Extraction Department

[0155] 12 Data Processing Department

[0156] 14 Container Cargo Information

[0157] 15 Unloading Container Information

[0158] 16 Container Storage Information

[0159] 20 container stacking planning device

[0160] 21 outbound prediction model

[0161] 22 Container Stacking Plan Model

[0162] 23 Restrictions

[0163] 24 evaluation indicators

[0164] Substrate analysis auxiliary system for 30A and 30B terminal devices.

Claims

1. A container stacking planning device for formulating a container stacking plan based on the conditions of a container terminal, characterized in that, include: The shipping sequence prediction unit takes as input information the attributes of containers that have been unloaded to the container terminal and shipped in the past, namely container cargo information, and the actual shipping date of each of the containers shipped in the past, and generates a prediction model for predicting the shipping date of each container. The unit takes as input the container cargo information attached to the containers unloaded to the container terminal and the containers stacked at the container terminal the shipping sequence information of the container cargo attached to the prediction model to predict the shipping date of each container and assign a shipping sequence to each container. and The container storage planning department takes as input the container cargo information attached to the containers unloaded at the container terminal, information indicating the storage locations of containers in the container terminal (i.e., storage container information), and the shipping priority order of each container calculated by the shipping sequence prediction department. Based on the constraints that should be considered during container loading and unloading operations, and in a manner that satisfies evaluation indicators used to evaluate container storage plans that are conducive to loading and unloading operations, it determines the storage location of each container in the container terminal. The evaluation criteria include a warehouse transfer evaluation condition that evaluates the storage location in a way that minimizes the number of containers stacked on top of containers with higher shipping priority as predicted by the shipping priority prediction unit, compared to containers with lower shipping priority. The limiting conditions include several unobstructed stacking conditions: when stacking containers at a container terminal, the stacked container configuration does not obstruct the container terminal's loading and unloading equipment from stacking and retrieving containers. For the aforementioned warehouse transfer evaluation conditions and the multiple barrier-free stacking conditions, weighting can be set according to which condition is emphasized when formulating a container stacking plan.

2. The container stacking planning device as described in claim 1, characterized in that: The restrictions also include the following condition: for a shell that constitutes the smallest unit zone in a container terminal for storing a specified number of containers, containers with the same specified attributes are stored in the same shell.

3. The container stacking planning device as described in claim 1, characterized in that: The container stacking planning department, for each container to be stacked, selects, according to the aforementioned constraints, the shells that best meet the evaluation criteria from the smallest unit divisions in the container terminal used for stacking a specified number of containers.

4. The container stacking planning device as described in claim 3, characterized in that: The container storage planning department selects the storage location that best meets the evaluation criteria, provided that containers are already stored in the selected container according to the constraints.

5. The container stacking planning device as described in claim 4, characterized in that: The evaluation index includes the following: minimizing the number of containers stacked on top of containers with higher shipping priority compared to containers with lower shipping priority.

6. The container stacking planning device as described in claim 1, characterized in that: The shipment sequence prediction unit compares the predicted shipment date of each container, which is based on the past container cargo information, with the actual shipment date recorded in the past container cargo information to generate the prediction model used in the shipment sequence prediction process in a way that minimizes the difference in the comparison.

7. A container stacking planning system for developing and applying a container stacking plan based on the conditions of a container terminal, characterized in that: Including container stacking planning equipment and terminal equipment, The container stacking planning device mentioned above includes: The container terminal business system stores: information representing the attributes of containers that have been unloaded to the container terminal and shipped out in the past, i.e., container cargo information; information including the unloading sequence of unloaded containers, i.e. unloading container information; and information representing the storage location of containers stored at the container terminal, i.e., storage container information. The shipping sequence prediction unit takes as input the container cargo information of containers that have been unloaded at the container terminal and shipped out in the past, and the actual shipping dates of each of the previously shipped containers, to generate a prediction model for predicting the shipping dates of each container. It takes as input the container cargo information of containers unloaded at the container terminal and containers stored at the container terminal the prediction model to predict the shipping dates of each container and assign a shipping sequence to each container; and The container storage planning department takes the unloading container information, the storage container information, and the shipping priority order of each container calculated by the shipping sequence prediction department as inputs. Based on the constraints that should be considered during container loading and unloading operations, and in a manner that satisfies the evaluation indicators used to evaluate container storage plans that are conducive to loading and unloading operations, it determines the storage location of each container in the container terminal. The terminal device receives and displays the container storage plan generated by the container storage planning device. The evaluation criteria include a warehouse transfer evaluation condition that evaluates the storage location in a way that minimizes the number of containers stacked on top of containers with higher shipping priority as predicted by the shipping priority prediction unit, compared to containers with lower shipping priority. The limiting conditions include several unobstructed stacking conditions: when stacking containers at a container terminal, the stacked container configuration does not obstruct the container terminal's loading and unloading equipment from stacking and retrieving containers. For the aforementioned warehouse transfer evaluation conditions and the multiple barrier-free stacking conditions, weighting can be set according to which condition is emphasized when formulating a container stacking plan.

8. The container stacking planning system as described in claim 7, characterized in that: The container stacking planning department, for each container to be stacked, selects, according to the aforementioned constraints, the shells that best meet the evaluation criteria from the smallest unit divisions in the container terminal used for stacking a specified number of containers.

9. The container stacking planning system as described in claim 8, characterized in that: The container storage planning department selects the storage location that best meets the evaluation criteria, provided that containers are already stored in the selected container according to the constraints.

10. The container stacking planning system as described in claim 9, characterized in that: The evaluation index includes the following: minimizing the number of containers stacked on top of containers with higher shipping priority compared to containers with lower shipping priority.

11. A container stacking planning method for developing a container stacking plan based on the conditions of a container terminal, characterized in that, An information processing apparatus having an arithmetic unit and a storage unit performs the following steps: The process involves taking information representing the attributes of containers that have been unloaded at the container terminal and shipped out in the past, i.e., container cargo information, and the actual shipping date of each of the containers that have been shipped out in the past, as inputs to generate a prediction model for predicting the shipping date of each container; taking the container cargo information attached to the containers unloaded at the container terminal and the containers stacked at the container terminal as inputs to the prediction model to predict the shipping date of each container and assign a shipping order to each container. and The process involves taking as input the container cargo information attached to the containers unloaded at the container terminal, information indicating the container storage locations within the terminal (i.e., container storage information), and the shipping priority order of each container calculated by the shipping sequence prediction unit. Based on constraints that must be considered during container loading and unloading operations, and in a manner that satisfies evaluation indicators used to assess container storage plans that are conducive to loading and unloading operations, the process determines the storage location of each container within the container terminal. The evaluation criteria include a warehouse transfer evaluation condition that evaluates the storage location in a way that minimizes the number of containers stacked on top of containers with higher shipping priority as predicted by the shipping priority prediction unit, compared to containers with lower shipping priority. The limiting conditions include several unobstructed stacking conditions: when stacking containers at a container terminal, the stacked container configuration does not obstruct the container terminal's loading and unloading equipment from stacking and retrieving containers. For the aforementioned warehouse transfer evaluation conditions and the multiple barrier-free stacking conditions, weighting can be set according to which condition is emphasized when formulating a container stacking plan.

12. The container stacking planning method as described in claim 11, characterized in that: The information processing device, for each container to be stacked, selects, according to the aforementioned constraints, the shell that best meets the evaluation criteria from the shells constituting the smallest unit area in the container terminal used for stacking a specified number of containers.

13. The container stacking planning method as described in claim 12, characterized in that: The information processing device, for each container to be stored, selects the storage location that best meets the evaluation criteria, provided that the containers are already stored in the selected container according to the constraints.

14. The container stacking planning method as described in claim 13, characterized in that: The evaluation index includes the following: minimizing the number of containers stacked on top of containers with higher shipping priority compared to containers with lower shipping priority.

Images