Stockpiling strategy optimization method for general goods wharf typical operation scene

Abstract

The invention discloses a general-purpose general cargo wharf typical operation scene stockpiling strategy optimization method, and relates to the technical field of port logistics management, and the method comprises the steps: generating a storage yard global state true phase source data pool through multi-source data perception and fusion processing; carrying out storage yard state entropy calculation and chaos degree quantitative evaluation based on the storage yard global state true phase source data pool to obtain global and partitioned real-time entropy monitoring reports; according to global and partitioned real-time entropy monitoring reports, carrying out adaptive switching decision making to obtain a strategy switching instruction and a new strategy parameter; and according to the strategy switching instruction and the new strategy parameter, carrying out dynamic reconstruction and elastic resource allocation on the virtual stockpiling container to obtain a latest virtual stockpiling container distribution diagram. According to the method, collaborative stockpiling plan generation and operation instruction issuing are performed based on the newest distribution diagram to obtain a plan diagram and an instruction list, decision rules are optimized and updated through closed-loop feedback and entropy learning after execution and monitoring, and self-adaptive optimization of a stockpiling strategy is realized.

Description

Technical Field

[0001] This invention relates to the field of port logistics management technology, and in particular to a method for optimizing storage strategies in typical operational scenarios at general cargo terminals. Background Technology

[0002] In the operation and management of general cargo terminals, optimizing yard storage strategies is a key aspect of improving overall operational efficiency and capabilities. Existing technologies predefine various storage strategies and select preset strategies for different operational scenarios based on static data such as ship arrival schedules and container attributes in the terminal's operating system. This approach, by establishing a strategy library, achieves a certain degree of responsiveness to diverse operational needs. Its decision-making process relies on summarizing and solidifying historical experience rules, aiming to maintain yard operational order through rule matching. This represents an important manifestation of the current intelligent development of automated terminals.

[0003] However, the management method based on the fixed rule strategy library mentioned above relies on making decisions based on pre-set, relatively static rules. When faced with the inherent diversity of operation scenarios in general cargo terminals, dynamic fluctuations in ship arrival times, and real-time changes in mechanical status, the adaptability of this technology is limited. Its main shortcoming is that the decision-making mechanism lacks dynamic resilience and cannot quantitatively perceive and evaluate the real-time operating status of the yard. Therefore, it is difficult to adaptively trigger strategy switching based on changes in the actual disorder of the yard. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for optimizing the stacking strategy in typical operating scenarios of general cargo terminals, which solves the problem that existing technologies cannot dynamically switch adaptive strategies based on the real-time quantification of the disorder in the yard.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for optimizing the stacking strategy in typical operation scenarios of general cargo terminals, which includes generating a global state truth source data pool of the yard through multi-source data perception and fusion processing.

[0008] Based on the source data pool of the global state of the yard, the entropy value of the yard state is calculated and the disorder is quantitatively assessed to obtain real-time entropy value monitoring reports for the global and regional levels.

[0009] Adaptive switching decisions are made based on real-time entropy monitoring reports of the global and partition levels, resulting in policy switching instructions and new policy parameters;

[0010] Based on the policy switching instructions and new policy parameters, the virtual heap containers are dynamically reconstructed and elastically allocated resources to obtain the latest virtual heap container distribution map.

[0011] Based on the latest virtual storage container distribution map, collaborative storage plan generation and operation instruction issuance are performed to obtain the storage plan map and operation instruction list.

[0012] The system executes and monitors the stacking plan and work instruction list to obtain real-time work execution data and updated stacking status. Based on the real-time work execution data and updated stacking status, it performs closed-loop feedback and entropy learning optimization to obtain updated decision rules and policy switching instructions.

[0013] As a preferred embodiment of the general cargo terminal typical operation scenario stacking strategy optimization method described in this invention, the method includes the following steps: generating a global state truth source data pool for the yard through multi-source data perception and fusion processing:

[0014] Collect container attribute information, real-time location and status of yard machinery, ship arrival forecasts and historical operation data, and transmit them to the central data unit to remove obviously invalid values ​​and correct identifiable erroneous formats.

[0015] Based on timestamps and container numbers, the system processes container attribute information, real-time location and status of yard machinery, ship arrival forecasts, and historical operation data to generate a global yard status source data pool.

[0016] As a preferred embodiment of the general cargo terminal typical operation scenario stacking strategy optimization method described in this invention, wherein: a distributed parallel data query method is used to extract the destination port code, container size code and operation status identifier of all containers in each stacking unit in the global state truth source data pool of the yard, forming a unit container attribute list;

[0017] Based on the list of unit container attributes, the destination port entropy value of each storage unit is calculated using the information entropy formula.

[0018] Based on the container attribute list of the stacking unit, the container type entropy value of each stacking unit is calculated using the information entropy formula;

[0019] Based on the container attribute list of the stacking unit, the operational state entropy value of each stacking unit is calculated using the information entropy formula;

[0020] Based on the destination port entropy value, container entropy value, and operational status entropy value of each storage unit, the disorder index of each storage unit is calculated using a weighted summation method.

[0021] By integrating the destination port entropy value, container entropy value, state entropy value, and disorder index, a real-time entropy monitoring report is obtained for both global and regional levels.

[0022] As a preferred embodiment of the storage strategy optimization method for a typical operational scenario of a general cargo terminal as described in this invention, the method includes the following steps: making adaptive switching decisions based on real-time entropy monitoring reports of the global and regional data to obtain strategy switching instructions and new strategy parameters.

[0023] Read the global average comprehensive disorder index, the regional average comprehensive disorder index, and the list of high disorder hotspot areas recorded in the global and regional real-time entropy monitoring reports to obtain the current overall and local disorder status of the storage yard;

[0024] In the configuration parameters of the heap strategy mode library, preset the global entropy threshold and the partition entropy threshold. Compare the global average comprehensive disorder index with the global entropy threshold. If the global average comprehensive disorder index exceeds the global entropy threshold, the global state is determined to be abnormal. Compare the partition average comprehensive disorder index with the partition entropy threshold. If the partition average comprehensive disorder index exceeds the threshold, the partition state is determined to be abnormal.

[0025] Based on the global state anomaly determination results and the partition state anomaly determination results, select the heap strategy mode for the current chaotic state from the heap strategy mode library.

[0026] Based on the heap storage strategy mode, generate strategy switching instructions and new strategy parameters.

[0027] As a preferred embodiment of the storage strategy optimization method for a typical operational scenario of a general cargo terminal as described in this invention, the method involves: dynamically reconstructing and elastically allocating virtual storage containers based on strategy switching instructions and new strategy parameters to obtain the latest virtual storage container distribution map, including the following steps:

[0028] Semantic parsing is used to parse the policy switching command and new policy parameters to obtain the new policy mode identifier and configuration parameters.

[0029] Based on the new strategy mode identifier and configuration parameters, create a new virtual heap container at the logical level;

[0030] Based on the new strategy mode identifier and configuration parameters, adjust the attributes of the virtual storage container, adjust the allowed container type, maximum stacking height and associated flight number for operating the virtual storage container, and obtain the adjusted virtual storage container information;

[0031] Based on the new strategy mode identifier and corresponding configuration parameters, the physical storage space occupied by the disbanded virtual storage container is remarked as available and included in the allocable resource set. The virtual storage containers that have completed the operation voyage are disbanded to obtain the retained virtual storage container information.

[0032] Integrate the information of newly created, adjusted, and retained virtual heap containers to generate the latest virtual heap container distribution map.

[0033] As a preferred embodiment of the method for optimizing the stacking strategy in a typical operational scenario of a general cargo terminal as described in this invention, the method involves: generating a collaborative stacking plan and issuing operational instructions based on the latest virtual stacking container distribution map to obtain a stacking plan map and a list of operational instructions, including the following steps:

[0034] The structured data parsing interface is used to read the spatial boundaries, attribute parameters and status information of each virtual storage container marked in the latest virtual storage container distribution map, obtain the constraints of the currently available storage resources, and extract the detailed information of the containers to be operated from the source data pool of the global status of the yard.

[0035] Based on the detailed information of the containers to be operated and the latest virtual storage container distribution map, an optimized allocation algorithm is used to match the virtual storage container position for each container to be operated, and the matching result is obtained.

[0036] Based on the matching results, a detailed visual stacking plan is generated, which indicates the stacking position and stacking order of each container to be processed in the virtual stacking container;

[0037] A specific list of job instructions is generated based on the heaping plan diagram.

[0038] As a preferred embodiment of the general cargo terminal typical operation scenario stacking strategy optimization method described in this invention, the method includes the following steps: executing the stacking plan and operation instruction list and monitoring them to obtain real-time operation execution data and updated yard status:

[0039] The stacking plan and work instruction list are sent to the yard crane operation terminal and the horizontal transport vehicle dispatching terminal.

[0040] The actual container pickup location coordinates, container drop location coordinates, and operation timestamps are collected from the yard crane operation terminal by IoT devices to form yard crane operation execution data records. The container transfer start point coordinates, end point coordinates, and transportation route trajectory are collected from the horizontal transport vehicle dispatch terminal to form horizontal transport execution data records.

[0041] Integrate the data records of yard crane operation and horizontal transportation to generate a real-time operation execution data summary table;

[0042] The real-time operation execution data summary table is used to update the true source data pool of the yard global status. Based on the updated true source data pool of the yard global status, the real-time operation execution data and the updated yard status are obtained.

[0043] As a preferred embodiment of the general cargo terminal typical operation scenario stacking strategy optimization method described in this invention, the following steps are included: Based on real-time operation execution data and the updated yard status, closed-loop feedback and entropy learning optimization are performed to obtain updated decision rules and strategy switching instructions:

[0044] Performance metrics are extracted from real-time operation execution data and updated yard status to form raw data for strategy effectiveness evaluation;

[0045] The raw data for evaluating the effectiveness of the strategy are correlated with the strategy mode identifier and the entropy value monitoring report data at the trigger time in the historical strategy switching instruction records. The correlated data are then classified and integrated according to the strategy mode to construct a sample dataset of strategy effectiveness.

[0046] Based on the policy effect sample dataset, the random forest regression algorithm is used to train the mapping relationship between policy switching instructions and execution effects. The entropy learning optimization model parameters are analyzed by different policy modes under different initial heap conditions to obtain the trained entropy learning optimization model.

[0047] The updated decision rules are generated by optimizing the model through entropy learning. The updated decision rules are then output to the policy switching instruction generation process, where they replace the entries in the original rule base.

[0048] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the stacking strategy optimization method for typical operating scenarios of general cargo terminals as described in the first aspect of the present invention.

[0049] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the stacking strategy optimization method for typical operating scenarios of general cargo terminals as described in the first aspect of the present invention.

[0050] The beneficial effects of this invention are as follows: a global state source data pool for the storage yard is generated through multi-source data perception and fusion processing; based on the global state source data pool, the storage yard state entropy value is calculated and the disorder metric is evaluated to obtain a real-time entropy value monitoring report; adaptive switching decisions are made to obtain strategy switching instructions and new strategy parameters; then, the virtual storage container is dynamically reconstructed and elastically allocated to generate the latest distribution map; based on the latest distribution map, collaborative storage plan generation and operation instructions are issued to obtain a plan map and instruction list; after execution and monitoring, the decision rules are optimized and updated through closed-loop feedback and entropy learning to achieve adaptive optimization of the storage strategy. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 A flowchart illustrating the optimization method for storage strategies in typical operational scenarios at a general cargo terminal.

[0053] Figure 2 This is a flowchart of the latest virtual heap container distribution diagram.

[0054] Figure 3 This is a schematic diagram of the source data pool for the global status of the storage yard.

[0055] Figure 4 Flowchart for Measurable Evaluation of Chaos Detailed Implementation

[0056] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0057] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0058] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0059] Reference Figures 1-4 As an embodiment of the present invention, this embodiment provides a method for optimizing the stacking strategy in a typical operation scenario of a general cargo terminal, including the following steps:

[0060] S1. Generate a global state source data pool for the storage yard through multi-source data perception and fusion processing.

[0061] S1.1 Collect container attribute information, real-time location and status of yard machinery, ship arrival forecasts and historical operation data, and transmit them to the central data unit to remove obviously invalid values ​​and correct identifiable erroneous formats;

[0062] Furthermore, through sensing devices and data interfaces deployed at the terminal site, the system continuously acquires container attribute information, including data such as size, container type, weight, and destination port; real-time location and status of yard machinery, including data such as crane coordinates, working status, and fault codes; ship arrival forecasts, including ship name, voyage number, and estimated berthing time; and historical operation data, including records of containers that have completed loading and unloading operations and equipment operation logs. All collected data is transmitted to the central data unit via wired or wireless networks for preliminary processing. The processing includes using a rule engine to remove non-numeric characters or outliers that are clearly outside the reasonable range in the weight field, and standardizing the date and time fields according to predefined format specifications to correct identifiable errors.

[0063] S1.2. Based on timestamps and container numbers, process container attribute information, real-time location and status of yard machinery, ship arrival forecasts and historical operation data to generate a global yard status source data pool.

[0064] Furthermore, based on a unified timestamp sequence and unique identifier for container number, the system performs correlation matching and alignment operations on the container attribute information, real-time location and status of yard machinery, ship arrival forecasts, and historical operation data that have completed preliminary processing. A time-series database is used to store and index the timestamped data streams, and various types of information involving the same container number within the same time window are integrated through correlation queries to generate a global status source data pool for the yard that includes multi-dimensional correlation information such as container physical location, machinery operation status, ship scheduling plans, and historical operation records.

[0065] S2. Based on the source data pool of the global state of the storage yard, calculate the entropy value of the storage yard state and conduct quantitative assessment of the disorder, and obtain real-time entropy monitoring reports for the global and regional states.

[0066] S2.1 Use a distributed parallel data query method to extract the destination port code, container size code and operation status identifier of all containers in each storage unit in the global status source data pool of the yard, and form a unit container attribute list.

[0067] Furthermore, a distributed parallel data query method is used to access the global state source data pool of the yard. Taking each storage unit as the basic unit, the destination port code field, container size code field, and operation status identifier field are retrieved and extracted in batches from all container records within the unit. The extracted three types of field data are organized and stored according to storage units to form a structured unit container attribute list, thus completing the aggregation of storage unit-level attribute data.

[0068] S2.2. Based on the list of unit container attributes, calculate the destination port entropy value of each storage unit using the information entropy formula;

[0069] The expression for the port of destination entropy is:

[0070] H d -∑(p i *log2(p i ));

[0071] Among them, H d p is the entropy value of the destination port. i Let i represent the percentage of containers at the i-th destination port, where i is the container index.

[0072] Furthermore, based on the list of unit container attributes, the destination port code set of each stacking unit is traversed, the frequency of different destination port codes appearing in the stacking unit is counted, the proportion of the number of containers at each destination port to the total number of containers in the unit is calculated, i.e., the probability value, and all probability values ​​are substituted into the information entropy formula to calculate the destination port entropy value of each stacking unit, thus quantifying the dispersion of the destination port distribution of containers in the unit.

[0073] S2.3 Based on the container attribute list of the stacking unit, calculate the container type entropy value of each stacking unit using the information entropy formula.

[0074] The expression for box entropy is:

[0075] H t ype=-∑(p j *log2(p j ));

[0076] Among them, H t ype is the box entropy value, p j Let j represent the percentage of containers of type j, where j is the container type index.

[0077] Furthermore, based on the unit container attribute list, the set of container size codes for each stacking unit is traversed, the frequency of different container size codes appearing in the unit is counted, the proportion of each container type to the total number of containers in the unit is calculated (i.e., the probability value), and all probability values ​​are substituted into the information entropy formula to calculate the container entropy value of each stacking unit, thus quantifying the uniformity of the mixing of different container sizes in the unit.

[0078] S2.4. Based on the container attribute list of the stacking unit, calculate the operational state entropy value of each stacking unit using the information entropy formula.

[0079] The expression for the job state entropy value is:

[0080] H s tatus=-[pm*log2([pm)+px*log2(px)];

[0081] Among them, Hs tatus is the entropy value of the operation status, pm is the proportion of imported containers, and px is the proportion of exported containers.

[0082] Furthermore, based on the unit container attribute list, the operation status identifier set of each stacking unit is traversed, the number of containers identified as imports and exports is counted respectively, the proportion of import containers and the proportion of export containers are calculated, and the two probability values ​​are substituted into the information entropy formula to calculate the operation status entropy value of each stacking unit, quantifying the degree of disorder of the mixing of import and export containers in the unit.

[0083] S2.5. Based on the destination port entropy value, container entropy value, and operational status entropy value of each storage unit, calculate the disorder index of each storage unit using a weighted summation method.

[0084] The expression for the disorder index is:

[0085] C = w d *H d +w t *H t ype+w s *H s tatus;

[0086] Where C is the disorder index, w d w is the weight of the destination port entropy value. t The weights for the box entropy values, w s This is the state entropy value.

[0087] Furthermore, based on the destination port entropy value, container entropy value, and operational status entropy value already calculated for each storage unit, a weighted summation method is used to calculate the comprehensive disorder index for each storage unit. This index comprehensively represents the overall disorder level of each storage unit.

[0088] S2.6 Integrate the destination port entropy value, container entropy value, state entropy value, and disorder index to obtain real-time entropy monitoring reports for the global and regional levels.

[0089] Furthermore, the destination port entropy value, container entropy value, operational status entropy value, and disorder index obtained from all storage units are integrated according to the storage unit number. Based on the predefined physical or logical partitions of the yard, the average entropy index of all units in the partition is calculated to generate a global and partition real-time entropy monitoring report that includes an overview of global entropy indexes, details of entropy indexes for each partition, and a list of high disorder units.

[0090] S3. Make adaptive switching decisions based on real-time entropy monitoring reports of the global and partition levels, and obtain policy switching instructions and new policy parameters.

[0091] S3.1 Read the global average comprehensive disorder index, the regional average comprehensive disorder index, and the list of high disorder hotspot areas recorded in the global and regional real-time entropy monitoring reports to obtain the current overall and local disorder status of the storage yard.

[0092] Furthermore, the system accesses and parses the latest generated global and partition real-time entropy monitoring reports, extracting the global average comprehensive disorder index, the average comprehensive disorder index of each partition, and the list of high disorder hotspots recorded in the reports. By reading these key indicators, the system can comprehensively understand the current disorder status of the storage yard at both the overall and partition levels.

[0093] S3.2. In the configuration parameters of the heap strategy mode library, preset the global entropy threshold and the partition entropy threshold. Compare the global average comprehensive disorder index with the global entropy threshold. If the global average comprehensive disorder index exceeds the global entropy threshold, the global state is determined to be abnormal. Compare the partition average comprehensive disorder index with the partition entropy threshold. If the partition average comprehensive disorder index exceeds the threshold, the partition state is determined to be abnormal.

[0094] Furthermore, the system queries the global entropy threshold and partition entropy threshold settings in the configuration parameters of the heap strategy mode library. It then compares the global average comprehensive disorder index value provided in the real-time entropy monitoring report with the global entropy threshold. When the global average comprehensive disorder index value is greater than the global entropy threshold, a global state anomaly determination result is generated. At the same time, the average comprehensive disorder index value of each partition is compared with the partition entropy threshold. When the average comprehensive disorder index value of any partition is greater than the partition entropy threshold, a state anomaly determination result is generated for that partition.

[0095] S3.3. Based on the global state anomaly determination results and the partition state anomaly determination results, select the heap storage strategy mode for the current chaotic state from the heap storage strategy mode library.

[0096] Furthermore, based on the severity and scope of the chaos indicated by the global state anomaly determination results and the partition state anomaly determination results, the most suitable heap strategy mode for dealing with the current chaotic state is matched and selected from the heap strategy mode library. The selection process prioritizes strategy modes that can effectively reduce the chaos of the identified abnormal areas. When multiple partitions experience state anomalies at the same time, the cruise convergence mode is selected to prioritize ensuring the order of critical operation channels.

[0097] S3.4. Generate policy switching instructions and new policy parameters according to the heap storage policy mode.

[0098] Furthermore, based on the specific requirements of the selected heap storage strategy mode, a strategy switching instruction containing a clear mode identifier is generated, and a new set of strategy parameters matching the mode is generated. The new set of strategy parameters includes specific configuration parameters such as dedicated heap storage area division rules, allowed heap storage box type restrictions, and maximum heap layer height restrictions.

[0099] S4. Based on the policy switching instructions and new policy parameters, dynamically reconstruct and elastically allocate resources to the virtual heap containers to obtain the latest virtual heap container distribution map.

[0100] S4.1 Use semantic parsing to parse the policy switching command and new policy parameters to obtain the new policy mode identifier and configuration parameters.

[0101] Furthermore, semantic parsing is used to process the received policy switching instructions and new policy parameters. By identifying keywords and parameter structures in the instruction text, the new policy mode identifier and corresponding configuration parameter details are accurately extracted, clarifying the policy type and specific operation requirements to be executed.

[0102] S4.2. Based on the new strategy mode identifier and configuration parameters, create a new virtual heap container at the logical level.

[0103] Furthermore, based on the new strategy mode identifier and configuration parameters, when creating a new virtual storage container at the logical level, the new strategy mode identifier is parsed to determine the type of virtual storage container to be created. Based on the voyage convergence mode, a dedicated voyage container needs to be created. Subsequently, according to the boundary rules, attribute constraints, and resource allocation requirements defined in the configuration parameters, the space range of virtual storage containers that meet the new strategy objectives is divided in the yard management logical space. The virtual storage container is then assigned attribute parameters such as a dedicated voyage identifier, a list of allowed storage container types, and a maximum stacking height limit specified by the configuration parameters, thus completing the logical creation of a virtual storage container that meets the requirements of the new strategy.

[0104] S4.3. Based on the new strategy mode identifier and configuration parameters, adjust the attributes of the virtual storage container, adjust the allowed container type, maximum stacking height and associated flight number for operating the virtual storage container, and obtain the adjusted virtual storage container information.

[0105] Furthermore, based on the requirements of the new strategy mode identifier and configuration parameters, the attributes of existing virtual storage containers in the yard are dynamically modified. The adjustment operation specifically involves updating the list of allowed container types for virtual storage containers, modifying the maximum stacking height limit of virtual storage containers, rebinding the associated voyage list of virtual storage containers, and generating adjusted virtual storage container information that matches the current strategy requirements.

[0106] S4.4. Based on the new strategy mode identifier and corresponding configuration parameters, the physical storage space occupied by the disbanded virtual storage container is remarked as available and included in the allocable resource set. The virtual storage containers that have completed the operation voyage are disbanded to obtain the retained virtual storage container information.

[0107] Furthermore, based on the new strategy mode identifier and corresponding configuration parameters, identify and disband virtual storage containers that have completed operational voyages or no longer meet the requirements of the new strategy, remark the physical storage space occupied by the virtual storage containers as available, and re-incorporate the released space resources into the allocable resource set for unified management, and filter and retain the information of virtual storage containers that still meet the policy requirements.

[0108] S4.5 Integrate the information of newly created, adjusted and retained virtual heap containers to generate the latest virtual heap container distribution map.

[0109] Furthermore, the newly created virtual storage container information, the virtual storage container information with adjusted attributes, and the selected and retained virtual storage container information are merged and integrated, and arranged and combined according to spatial location relationships to generate a complete and accurate latest virtual storage container distribution map that reflects the current logical space division of the storage yard.

[0110] S5. Based on the latest virtual storage container distribution map, generate collaborative storage plans and issue work instructions to obtain storage plan map and work instruction list.

[0111] S5.1 Use the structured data parsing interface to read the spatial boundaries, attribute parameters and status information of each virtual storage container marked in the latest virtual storage container distribution map, obtain the currently available storage resource constraints, and extract the detailed information of the containers to be operated from the global status source data pool of the yard.

[0112] Furthermore, by calling the structured data parsing interface to access the latest virtual storage container distribution map, the spatial boundary coordinate data, attribute parameter definitions including the allowed container type list and maximum stacking height limit, and status information including the current idle capacity contained in each virtual storage container element in the map are read and parsed. In this way, the constraints of the currently available storage resources are obtained, and detailed information of all containers to be operated, including container number, size, weight, destination port and operation priority, is retrieved and extracted from the global status truth source data pool of the yard.

[0113] S5.2 Based on the detailed information of the containers to be operated and the latest virtual storage container distribution map, an optimized allocation algorithm is used to match the virtual storage container position for each container to be operated, and the matching result is obtained.

[0114] Furthermore, based on the detailed information of the containers to be operated and the storage resource constraints provided by the latest virtual storage container distribution map, an optimization allocation algorithm is used to match containers with containers, satisfying the virtual storage container attribute parameters as hard constraints, in order to minimize the overall operation time and transportation distance. The optimal virtual storage container allocation position is calculated for each container to be operated, and the matching result includes the container number, the target virtual storage container number, and the specific bay layer number.

[0115] S5.3. Based on the matching results, generate a detailed visual stacking plan diagram. The stacking plan diagram indicates the stacking position and stacking order of each container to be operated in the virtual stacking container.

[0116] Furthermore, based on the matching results, a detailed visual stacking plan is generated. The stacking plan graphically displays the space division within each virtual stacking container and clearly marks the specific stacking position of each container to be operated within the target virtual stacking container, including bay position, layer height, and stacking order, using numbering labels and color blocks.

[0117] S5.4 Generate a specific list of job instructions based on the heaping plan diagram.

[0118] Furthermore, based on the completed stacking plan diagram, a specific list of operation instructions is generated. The list of operation instructions details the specific location coordinates and target coordinates of the container picking, placing, and stacking operations that the yard crane needs to perform for each operation step, as well as the container transfer start and end points and route planning that the horizontal transport vehicles need to perform, forming a set of operation commands that can be directly issued.

[0119] S6. Execute the stacking plan and work instruction list and monitor them to obtain real-time work execution data and updated stacking status.

[0120] S6.1. Distribute the stacking plan and work instruction list to the yard crane operation terminal and the horizontal transport vehicle dispatching terminal.

[0121] Furthermore, the generated stacking plan and work instruction list are transmitted to the yard crane operation terminal and the horizontal transport vehicle dispatching terminal via the communication network, respectively, to ensure that the terminal equipment receives complete visual stacking location guidance and executable operation command sequence, and starts the field equipment to execute the work process according to the plan.

[0122] S6.2. Collect the actual container pickup location coordinates, container drop location coordinates, and operation timestamps fed back by the yard crane operation terminal through IoT devices to form a yard crane operation execution data record. Collect the container transfer start point coordinates, end point coordinates, and transportation route trajectory fed back by the horizontal transport vehicle dispatch terminal to form a horizontal transport execution data record.

[0123] Furthermore, by deploying IoT sensors on the yard cranes, the actual container retrieval coordinates, actual container placement coordinates, and precise operation timestamps corresponding to each operation are collected in real time from the yard crane operation terminal. This data is then integrated in the order of operations to form a yard crane operation execution data record. At the same time, by using onboard IoT devices to collect the container transfer start-up coordinates, end-point coordinates, and trajectory point sequences of the actual driving path from the horizontal transport vehicle dispatch terminal, horizontal transport execution data records are formed, comprehensively capturing the spatiotemporal data during the operation execution process.

[0124] S6.3 Integrate the data records of yard bridge operation execution and horizontal transportation execution to generate a real-time operation execution data summary table.

[0125] Furthermore, the collected yard crane operation execution data records and horizontal transport execution data records are linked and aligned according to timestamps and container numbers, duplicate records are removed and missing fields are filled in, and all operation data is stored in a unified structured manner to generate a real-time operation execution data summary table containing complete information such as operation sequence, equipment actions, spatial displacement and time consumption.

[0126] S6.4 Update the true source data pool of the global status of the yard according to the real-time operation execution data summary table, and obtain the real-time operation execution data and the updated yard status based on the updated true source data pool of the global status of the yard.

[0127] Furthermore, based on the actual operation results recorded in the real-time operation execution data summary table, the yard global status source data pool is incrementally updated. The update operations include modifying the latest coordinate position of the moved containers, releasing the storage space occupancy status of completed operations, and marking the virtual storage container attributes of newly stacked containers to ensure that the data pool accurately reflects the latest physical status of the yard. Finally, based on the updated yard global status source data pool, real-time operation execution data and updated yard status reports are output.

[0128] S7. Based on real-time operation execution data and the updated yard status, perform closed-loop feedback and entropy learning optimization to obtain updated decision rules and strategy switching instructions.

[0129] S7.1 Extract performance indicators from real-time operation execution data and updated yard status to form raw data for strategy effectiveness evaluation.

[0130] Furthermore, key performance indicators are selected from real-time operation execution data and updated yard status, including operation completion time, actual number of container turning operations, equipment idle waiting time, and yard disorder change value. These indicators are organized by operation batch and marked with corresponding time windows to form raw data for strategy effect evaluation.

[0131] S7.2. Associate the raw data for strategy effect evaluation with the strategy mode identifier and entropy monitoring report data at the trigger time in the historical strategy switching instruction record. Then, classify and integrate the associated data according to the strategy mode to construct a strategy effect sample dataset.

[0132] Furthermore, the raw data for strategy effectiveness evaluation is matched and associated with historical strategy switching instruction records. By matching the strategy mode identifier and trigger timestamp in the strategy switching instruction records, the corresponding historical entropy monitoring report data is found. The strategy mode identifier, the entropy monitoring report data at the trigger timestamp, and the raw data for strategy effectiveness evaluation are integrated by job batch to form a strategy effectiveness sample dataset classified by strategy mode.

[0133] S7.3. Based on the policy effect sample dataset, the random forest regression algorithm is used to train the mapping relationship between policy switching instructions and execution effects. The entropy learning optimization model parameters are analyzed by different policy modes under different initial heap conditions to obtain the trained entropy learning optimization model.

[0134] Furthermore, based on the policy effect sample dataset, a random forest regression algorithm is used for training. During training, the policy mode identifier in historical policy switching instructions and the entropy monitoring report data at the trigger time are used as input features, and the performance indicators in the original policy effect evaluation data are used as target variables. The mapping relationship between input features and output effects is established through ensemble learning of multiple decision trees. The model parameters are optimized by analyzing the performance of each policy mode under different initial stockpile states, and finally, the trained entropy learning optimization model is obtained.

[0135] S7.4. The updated decision rules are generated by optimizing the model through entropy learning. The updated decision rules are output to the policy switching instruction generation process, and the updated decision rules will replace the entries in the original rule base.

[0136] Furthermore, the trained entropy learning optimization model generates updated decision rules, which include an optimized global entropy threshold setting scheme, partition entropy weight adjustment coefficients, and policy pattern matching priority sequence. These updated decision rules are output to the policy switching instruction generation process and replace the corresponding entries in the original rule base, thus completing the iterative optimization of the policy decision mechanism.

[0137] This embodiment also provides a computer device applicable to the stacking strategy optimization method for typical operation scenarios of general cargo terminals, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the stacking strategy optimization method for typical operation scenarios of general cargo terminals as proposed in the above embodiment.

[0138] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0139] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the stacking strategy optimization method for typical operational scenarios of general cargo terminals as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0140] In summary, this invention generates a global state truth source data pool for the storage yard through multi-source data perception and fusion processing. Based on this data pool, it calculates the storage yard state entropy and performs quantitative assessment of disorder to obtain a real-time entropy monitoring report. It then makes adaptive switching decisions to obtain strategy switching instructions and new strategy parameters. Subsequently, it dynamically reconstructs and elastically allocates resources to the virtual storage container to generate the latest distribution map. Based on the latest distribution map, it generates a collaborative storage plan and issues operation instructions to obtain a plan map and instruction list. After execution and monitoring, it optimizes and updates the decision rules through closed-loop feedback and entropy learning to achieve adaptive optimization of the storage strategy.

[0141] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing the stacking strategy in typical operational scenarios at a general cargo terminal, characterized by: Comprising, Through multi-source data perception and fusion processing, a global state truth source data pool of the yard is generated; Based on the global state truth source data pool of the yard, yard state entropy value calculation and chaos quantification evaluation are performed to obtain real-time entropy value monitoring reports of the global and partitions; According to the real-time entropy value monitoring reports of the global and partitions, adaptive switching decisions are made to obtain strategy switching instructions and new strategy parameters; According to the strategy switching instructions and new strategy parameters, dynamic reconstruction and elastic resource allocation are performed on the virtual storage containers to obtain the latest virtual storage container distribution map; Based on the latest virtual storage container distribution map, collaborative storage planning is generated and job instructions are issued to obtain a storage planning map and a job instruction list; The storage planning map and the job instruction list are executed and monitored to obtain real-time job execution data and updated yard states, and closed-loop feedback and entropy learning optimization are performed according to the real-time job execution data and the updated yard states to obtain updated decision rules and strategy switching instructions.

2. The general-purpose container terminal typical operation scenario stacking strategy optimization method according to claim 1, characterized in that: Through multi-source data perception and fusion processing, a global state truth source data pool of the yard is generated, including the following steps: Collect container attribute information, real-time position and state of yard machinery, ship arrival forecast and historical operation data, and transmit them to the central data processor for removing obvious invalid values and correcting identifiable error formats; Based on the timestamp and container number, the container attribute information, real-time position and state of yard machinery, ship arrival forecast and historical operation data are processed to generate the global state truth source data pool of the yard.

3. The method for optimizing the typical operation scenario stacking strategy of a general cargo terminal according to claim 2, characterized in that: Based on the global state truth source data pool of the yard, yard state entropy value calculation and chaos quantification evaluation are performed to obtain real-time entropy value monitoring reports of the global and partitions, including the following steps: Use a distributed parallel data query method to extract the destination port code, container type size code and job status identifier of all containers in each storage unit in the global state truth source data pool of the yard to form a unit container attribute list; Based on the unit container attribute list, calculate the destination port entropy value of each storage unit using the information entropy formula; Based on the unit container attribute list, calculate the container type entropy value of each storage unit using the information entropy formula; Based on the unit container attribute list, calculate the job status entropy value of each storage unit using the information entropy formula; Based on the destination port entropy value, container type entropy value and job status entropy value of each storage unit, calculate the chaos index of each storage unit using the weighted summation method; Integrate the destination port entropy value, container type entropy value, state entropy value and chaos index to obtain real-time entropy value monitoring reports of the global and partitions.

4. The general-purpose container terminal typical operation scenario stacking strategy optimization method according to claim 3, characterized in that: According to the real-time entropy value monitoring reports of the global and partitions, adaptive switching decisions are made to obtain strategy switching instructions and new strategy parameters, including the following steps: Read the global average comprehensive chaos index, partition average comprehensive chaos index and high chaos hotspot area list information recorded in the global and partition real-time entropy value monitoring reports to obtain the overall and local chaos state of the yard; The preset global entropy threshold and the partition entropy threshold are set in the configuration parameters of the stacking strategy mode library, the global average comprehensive confusion index is compared with the global entropy threshold, and the global state is determined to be abnormal when the global average comprehensive confusion index exceeds the global entropy threshold; the partition average comprehensive confusion index is compared with the partition entropy threshold, and the partition state is determined to be abnormal when the partition average comprehensive confusion index exceeds the threshold; Based on the global state abnormality determination result and the partition state abnormality determination result, a stacking strategy mode for the current confusion state is selected from the stacking strategy mode library; According to the stacking strategy mode, a strategy switching instruction and new strategy parameters are generated.

5. The general-purpose container terminal typical operation scenario stacking strategy optimization method according to claim 4, characterized in that: According to the strategy switching instruction and the new strategy parameters, the virtual stacking container is dynamically reconstructed and the elastic resource is allocated, and the latest virtual stacking container distribution map is obtained, including the following steps: The semantic analysis method is used to analyze the strategy switching instruction and the new strategy parameters to obtain the new strategy mode identifier and the configuration parameters; According to the new strategy mode identifier and the configuration parameters, a new virtual stacking container is created at the logical level; According to the new strategy mode identifier and the configuration parameters, the virtual stacking container attributes are adjusted, the allowed box type, the maximum stacking height and the associated voyage of the virtual stacking container are adjusted, and the adjusted virtual stacking container information is obtained; According to the new strategy mode identifier and the corresponding configuration parameters, the physical stacking space occupied by the virtual stacking container is dissolved and marked as available, and is included in the allocatable resource set, the virtual stacking container of the completed work voyage is dissolved, and the retained virtual stacking container information is obtained; The new, adjusted and retained virtual stacking container information is integrated to generate the latest virtual stacking container distribution map.

6. The general-purpose container terminal typical operation scenario stacking strategy optimization method according to claim 5, characterized in that: Based on the latest virtual stacking container distribution map, a collaborative stacking plan is generated and a work instruction is issued, and a stacking plan map and a work instruction list are obtained, including the following steps: The structured data analysis interface is used to read the space boundary, attribute parameters and state information of each virtual stacking container marked in the latest virtual stacking container distribution map, to obtain the current available stacking resource constraint condition, and to extract the detailed information of the to-be-worked container from the yard global state truth source data pool; Based on the detailed information of the to-be-worked container and the latest virtual stacking container distribution map, an optimal allocation algorithm is used to match the virtual stacking container position for each to-be-worked container, and a matching result is obtained; According to the matching result, a detailed visual stacking plan map is generated, and the stacking plan map marks the stacking position and stacking sequence of each to-be-worked container in the virtual stacking container; Based on the stacking plan map, a specific work instruction list is generated.

7. The general cargo terminal typical operation scenario stacking strategy optimization method according to claim 6, characterized in that: The stacking plan map and the work instruction list are executed and monitored to obtain real-time work execution data and updated yard state, including the following steps: The stacking plan map and the work instruction list are issued to the yard crane terminal and the horizontal transportation vehicle scheduling terminal; The actual box picking position coordinates, box placing position coordinates and operation time stamps fed back by the yard crane terminal are collected through the Internet of Things equipment to form a yard crane operation execution data record, and the container transfer start point coordinates, end point coordinates and transportation path trajectory fed back by the horizontal transportation vehicle scheduling terminal are collected to form a horizontal transportation execution data record; Integrate the yard crane operation execution data record and the horizontal transportation execution data record to generate a real-time job execution data summary table; Update the yard global state truth source data pool according to the real-time job execution data summary table, and obtain real-time job execution data and an updated yard state based on the updated yard global state truth source data pool.

8. The general cargo terminal typical operation scenario stacking strategy optimization method according to claim 7, characterized in that, According to the real-time job execution data and the updated yard state, perform closed-loop feedback and entropy learning optimization to obtain updated decision rules and policy switching instructions, including the following steps: Extract performance indicators from the real-time job execution data and the updated yard state to form policy effect evaluation raw data; Correlate the policy effect evaluation raw data with policy mode identifiers and trigger time entropy value monitoring report data in the historical policy switching instruction record, integrate the correlated data according to policy modes, and construct a policy effect sample data set; Based on the policy effect sample data set, train a mapping relationship between policy switching instructions and execution effects by using a random forest regression algorithm, analyze entropy learning optimization model parameters for each policy mode under different initial yard states, and obtain a trained entropy learning optimization model; Generate updated decision rules by using the entropy learning optimization model, and output the updated decision rules to a policy switching instruction generation process. The updated decision rules will replace entries in an original rule library. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to implement the steps of the general cargo terminal typical job scene stacking strategy optimization method of any one of claims 1-8.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the general cargo terminal typical job scene stacking strategy optimization method of any one of claims 1-8.

Images