Edge cooperation-based logistics park platform intelligent management and control method and system

Abstract

The application discloses a logistics park platform intelligent management and control method and system based on edge cooperation, and relates to the technical field of park platform management. The method comprises the following steps: a target warehouse has multiple cargo storage areas and multiple platforms, and a distance matrix of the multiple cargo storage areas and the multiple platforms is established; vehicle position information and operation demand information of multiple transport vehicles are acquired in real time, and multiple vehicle arrival times are predicted and acquired according to the vehicle position information of the multiple transport vehicles; an initial vehicle platform matching scheme is generated according to the distance matrix and the operation demand information, conflict detection is performed on the initial vehicle platform matching scheme based on the multiple vehicle arrival times, and a conflict detection result is acquired; when there is a conflict vehicle platform pair in the conflict detection result, the initial vehicle platform matching scheme is optimized according to the distance matrix, an optimized vehicle platform matching scheme is generated, and logistics park platform management and control are performed. The application effectively improves scheduling accuracy and resource utilization efficiency.

Description

Technical Field

[0001] This invention relates to the field of park platform management technology, specifically to a method and system for intelligent control of logistics park platforms based on edge collaboration. Background Technology

[0002] With the rapid development of the logistics industry, logistics parks, as important hubs for cargo distribution and transshipment, have their platform management level directly affecting operational efficiency.

[0003] In existing technologies, logistics park platform management relies on manual scheduling or simple information management systems. Vehicles and logistics parks agree on approximate arrival times via telephone or a simple reservation system. When vehicles actually arrive at the park, a first-come, first-served queuing allocation model is adopted, that is, available platforms are allocated in the order of vehicle arrival.

[0004] Under this management method, vehicles may arrive early or late, making it impossible to plan platform resources in advance; the first-come, first-served allocation principle ignores the spatial layout inside the warehouse, which may assign vehicles that need to load and unload goods on the east side of the warehouse to the far west platform, increasing forklift transportation distance and operation time; when multiple vehicles arrive at the same time, there is a lack of intelligent allocation strategy, and it is impossible to optimize allocation based on factors such as operation type and distance.

[0005] The passive response management model in existing technologies cannot achieve proactive edge resource scheduling, resulting in low platform utilization, long vehicle waiting times, and low overall operational efficiency.

[0006] Therefore, existing technologies suffer from technical problems such as resource allocation conflicts and low scheduling efficiency caused by uncertain vehicle arrival times. Summary of the Invention

[0007] This invention provides a method and system for intelligent management and control of logistics park platforms based on edge collaboration, aiming to solve the technical problems of resource allocation conflicts and low scheduling efficiency caused by uncertain vehicle arrival times in the prior art.

[0008] In view of the above problems, the present invention provides a method and system for intelligent management and control of logistics park platforms based on edge collaboration.

[0009] In a first aspect, the present invention provides a method for intelligent management and control of logistics park platforms based on edge collaboration, including:

[0010] Identify the target warehouse, which has multiple cargo storage areas and multiple platforms, and establish a distance matrix between the multiple cargo storage areas and multiple platforms;

[0011] The system acquires real-time vehicle location information and operational requirements information of multiple transport vehicles, and predicts the arrival time of the multiple vehicles based on their location information.

[0012] An initial vehicle platform matching scheme is generated based on the distance matrix and the operation requirement information, and conflict detection is performed on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles to obtain the conflict detection results.

[0013] When conflicting vehicle-platform pairs are detected in the conflict detection results, the initial vehicle-platform matching scheme is optimized based on the distance matrix to generate an optimized vehicle-platform matching scheme for platform management in the logistics park.

[0014] Secondly, the present invention provides an intelligent management and control system for logistics park platforms based on edge collaboration, comprising:

[0015] A distance matrix establishment module is used to determine the target warehouse, which has multiple cargo storage areas and multiple platforms, and to establish a distance matrix between the multiple cargo storage areas and multiple platforms.

[0016] The arrival time prediction module is used to obtain the vehicle location information and operation demand information of multiple transport vehicles in real time, and predict the arrival time of the multiple vehicles based on the vehicle location information of the multiple transport vehicles.

[0017] The matching conflict detection module is used to generate an initial vehicle platform matching scheme based on the distance matrix and the operation requirement information, and to perform conflict detection on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles, and obtain the conflict detection results.

[0018] The matching scheme optimization module is used to optimize the initial vehicle platform matching scheme according to the distance matrix when there are conflicting vehicle platform pairs in the conflict detection results, and generate an optimized vehicle platform matching scheme for platform management in the logistics park.

[0019] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0020] This invention provides a method and system for intelligent management and control of logistics park platforms based on edge collaboration. By constructing a distance matrix between cargo storage areas and platforms, it provides precise spatial scheduling basis; accurately predicts vehicle arrival times, laying the timing foundation for matching and scheduling; identifies scheduling conflicts in advance through initial matching and conflict detection; and optimizes the matching scheme to resolve conflicts when they occur. This invention achieves the technical effect of intelligent allocation of multiple platforms based on edge collaborative prediction and two-dimensional spatiotemporal optimization, significantly improving scheduling accuracy and resource utilization efficiency. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0022] Figure 1 A flowchart illustrating the intelligent management and control method for logistics park platforms based on edge collaboration provided in an embodiment of the present invention;

[0023] Figure 2 A schematic diagram of the structure of the intelligent control system for logistics park platforms based on edge collaboration provided in an embodiment of the present invention;

[0024] The components represented by each number in the attached diagram are explained below:

[0025] 11 Distance matrix establishment module, 12 Arrival time prediction module, 13 Matching conflict detection module, 14 Matching scheme optimization module. Detailed Implementation

[0026] This invention provides a method and system for intelligent management and control of logistics park platforms based on edge collaboration, which is used to address the technical problems of resource allocation conflicts and low scheduling efficiency of multiple platforms caused by uncertain vehicle arrival times in the prior art.

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0028] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0029] Example 1, as Figure 1 As shown, this invention provides an intelligent management and control method for logistics park platforms based on edge collaboration, the method comprising:

[0030] S100: Determine the target warehouse, which has multiple cargo storage areas and multiple platforms, and establish a distance matrix between the multiple cargo storage areas and multiple platforms.

[0031] In this embodiment of the invention, a target warehouse is determined, which has multiple cargo storage areas and multiple platforms. A distance matrix is ​​established for these cargo storage areas and platforms. In the intelligent management and control process of logistics park platforms, the transportation distance between cargo storage areas and platforms is the core spatial basis for platform allocation. Under the traditional management and control model, staff rely heavily on manual experience to estimate distances without considering actual scenario factors such as the layout of the warehouse passageways. This leads to a large deviation between the estimated distance and the actual transportation distance, directly causing problems such as inaccurate vehicle transportation time predictions and frequent platform occupancy conflicts. Therefore, it is necessary to construct an accurate distance matrix through a standardized process to provide reliable spatial data support for subsequent vehicle-platform matching and conflict optimization.

[0032] Step S100 in the method provided in this embodiment of the invention includes:

[0033] This involves establishing a distance matrix for multiple cargo storage areas and multiple platforms, including:

[0034] Obtain a three-dimensional model of the target warehouse, determine the coordinate range of multiple cargo storage areas and the coordinate positions of multiple platforms in the three-dimensional model, and obtain multiple storage coordinate ranges and multiple platform coordinate positions;

[0035] The first platform is determined from the plurality of platforms, and its coordinate position is obtained;

[0036] Based on the channel layout in the three-dimensional model, calculate the shortest transportation path from the coordinate position of the first platform to each storage coordinate range, and obtain multiple transportation distance values ​​from the first platform to each cargo storage area.

[0037] By obtaining multiple transport distance values ​​from the first platform to each cargo storage area, multiple transport distance values ​​from the remaining platforms to each cargo storage area are obtained, resulting in multiple transport distance values ​​from the platforms to each cargo storage area.

[0038] The transportation distance values ​​from multiple platforms to each cargo storage area are aggregated to construct the distance matrix.

[0039] First, the target warehouse is identified, which has multiple cargo storage areas and multiple loading platforms. The target warehouse refers to a logistics warehouse covered by the control task that possesses cargo storage and loading / unloading functions; it serves as the spatial carrier for all subsequent calculations. For example, the target warehouse is warehouse A, containing three cargo storage areas A1, A2, and A3, and three loading platforms B1, B2, and B3. A1 is the fresh produce storage area, A2 is the daily necessities storage area, A3 is the home appliance storage area, B1 is platform number 1, located on the main entrance side, B1 is platform number 1, located in the middle of the warehouse, and B1 is platform number 1, located on the back door side.

[0040] Secondly, a 3D model of the target warehouse is obtained. Within this model, the coordinate ranges of multiple cargo storage areas and the coordinate positions of multiple platforms are determined, resulting in multiple storage coordinate ranges and multiple platform coordinate positions. The 3D model refers to a digital model of the warehouse constructed using BIM (Building Information Modeling) technology, containing 3D spatial data of all physical structures within the warehouse, including storage areas, platforms, transport channels, and columns. The storage coordinate range refers to the spatial coordinate interval of the cargo storage area within the 3D model, including X, Y, and Z axes, with the Z-axis representing height, and Z=0 at ground level. The platform coordinate positions are the positioning coordinates of the loading and unloading areas on the platforms. The midpoint of the platform edge is selected as the positioning point, with the Z-axis fixed at 0, flush with the ground. Autodesk Revit (BIM modeling software) was used to scan the physical space of warehouse A and generate a 3D model containing all structural details. The functional attributes of each area were marked in the model. A coordinate system was established: the southwest corner of the warehouse was the origin, the east direction was the positive X-axis, and the north direction was the positive Y-axis. In the coordinate system of the 3D model, the boundaries of each cargo storage area were delineated and the coordinate range was recorded. The midpoint of the loading and unloading edge of each platform was selected and the coordinate position was recorded.

[0041] For example, the 3D model of warehouse A includes three cargo storage areas A1, A2, and A3, and three platforms B1, B2, and B3. The coordinate ranges of the three cargo storage areas are determined as follows: A1: A2: A3: Determine the coordinates of the three platforms: B1: (X=62, Y=16, Z=0); B2: (X=38, Y=16, Z=0); B3: (X=62, Y=40, Z=0).

[0042] Subsequently, the first platform is determined from the multiple platforms, and its coordinates are obtained. The first platform refers to the first calculation object randomly selected from all platforms, used to establish a demonstration process for calculating the distance between the platform and the cargo storage area. Subsequent platforms are calculated using the same logic. One platform is randomly selected from all platforms in the target warehouse as the first platform; its coordinates are directly extracted from the constructed 3D model. For example, platform B1 is randomly selected from platforms B1, B2, and B3 in warehouse A as the first platform, with coordinates (X=62, Y=16, Z=0).

[0043] Furthermore, based on the channel layout in the 3D model, the shortest transportation path from the coordinate position of the first platform to each storage coordinate range is calculated, obtaining multiple transportation distance values ​​from the first platform to each cargo storage area. The channel layout refers to the pre-defined internal transportation channel network in the 3D model, including structural parameters such as main channels, branch channels, and turning areas. The shortest transportation path refers to the shortest feasible path from the platform coordinate position to the storage area coordinate range under the constraints of the channel layout, avoiding obstacles such as pillars and shelves, and meeting the requirements for vehicle passage. Channel layout data is exported from the 3D model, and a warehouse channel topology map is constructed, marking channel nodes, lengths, and directions. Using Dijkstra's algorithm (shortest path algorithm), with the first platform coordinates as the starting point and the center coordinates of the storage area coordinate range as the ending point (e.g., the center coordinates of A1 are (18, 14, 0), the shortest feasible path is obtained, and the calculated path length is the transportation distance value from the first platform to that storage area.

[0044] For example, in the topology diagram of warehouse A, the main channel C1 (X=12-62, Y=16, Z=0) connects B1 with each storage area, and the branch channel C11 (X=20, Y=16-22, Z=0) connects C1 with A1. Starting from B1 (62, 16, 0) and ending at the center of A1 (18, 14, 0), the path calculated using Dijkstra's algorithm is as follows: Starting from B1, travel along the main channel C1 for 42 meters, then along the branch channel C11 from Y=16 to Y=14 for 2 meters, for a total distance of 44 meters. Similarly, the path for B... The path from B1 to center A2 (38,14,0) is as follows: enter the main channel C1 from B1, travel along the main channel C1 from X=62 to X=38, with a distance of 24 meters; the path from B1 to center A3 (30,40,0) is as follows: enter the main channel C1 from B1, travel along the main channel C1 from X=62 to X=30, with a distance of 32 meters, then travel along the branch channel C12 from Y=16 to Y=40, with a distance of 24 meters. The total distance of the two routes is 56 meters; the final distances from B1 to each storage area are: A1=44m, A2=24m, A3=56m.

[0045] Furthermore, following the method of obtaining multiple transport distance values ​​from the first platform to each cargo storage area, multiple transport distance values ​​from the remaining platforms to each cargo storage area are obtained, resulting in multiple transport distance values ​​from each platform to each cargo storage area. Using the above calculation logic as a template, the second platform, third platform, and other remaining platforms are successively taken as starting points; the shortest transport distance value from each platform to all storage areas is calculated. For example, the final transport distance values ​​from the remaining platforms to each cargo storage area are: B2: A1=20m, A2=2m, A3=32m; B3: A1=68m, A2=52m, A3=32m.

[0046] Finally, the transportation distance values ​​from multiple platforms to each cargo storage area are summarized to construct the distance matrix. The distance matrix is ​​a two-dimensional data table with platforms as rows and cargo storage areas as columns. Each element in the matrix represents the shortest transportation distance from the platform in the corresponding row to the storage area in the corresponding column, serving as the data carrier for subsequent scheduling calculations. The row indices of the matrix are defined as all platforms in the target warehouse, and the column indices as all cargo storage areas, clarifying the correspondence between rows and columns. The calculated distance values ​​from each platform to each storage area are then associated one by one according to the relationship between the platform in the corresponding row and the storage area in the corresponding column, forming a complete distance matrix data. For example, the distance matrix of warehouse A has B1, B2, and B3 as rows and A1, A2, and A3 as columns. The distance from B1 to A1 is 44m, to A2 is 24m, and to A3 is 56m. The distance from B2 to A1 is 20m, to A2 is 2m, and to A3 is 32m. The distance from B3 to A1 is 68m, to A2 is 52m, and to A3 is 32m. These distances are then combined to form the distance matrix.

[0047] In this embodiment of the invention, by establishing a three-dimensional model and calculating the transportation distance from the platform to the storage area in conjunction with the channel layout, the problem of large deviations in traditional manual distance estimation is effectively solved. At the same time, the constructed distance matrix transforms the spatial relationship between the platform and the cargo storage area into structured data, which can be directly called by subsequent scheduling algorithms, avoiding errors in manual data entry. This provides accurate spatial data support for subsequent vehicle-platform matching and conflict optimization, laying the foundation for efficient scheduling.

[0048] S200: Real-time acquisition of vehicle location information and operation requirement information of multiple transport vehicles, and prediction of the arrival time of the multiple vehicles based on the vehicle location information of the multiple transport vehicles.

[0049] In this embodiment of the invention, the real-time location information and operational demand information of multiple transport vehicles are acquired, and the arrival times of these vehicles are predicted based on their location information. Vehicle arrival times are the timing basis for platform scheduling, directly determining the rationality of the matching scheme and the conflict rate. In traditional control models, arrival times are often directly estimated by navigation software, relying solely on real-time vehicle locations and road conditions, without considering individual factors such as driver habits and historical driving performance. This leads to high prediction errors, easily causing resource waste, and also depriving subsequent conflict detection of accurate timing support, resulting in scheduling scheme failure. Therefore, it is necessary to construct a precise prediction process with basic prediction and personalized correction to provide reliable timing data for intelligent platform control.

[0050] Step S200 in the method provided in this embodiment of the invention includes:

[0051] The method of predicting and obtaining the arrival times of the multiple transport vehicles based on their location information includes:

[0052] Obtain the warehouse location information of the target warehouse;

[0053] The vehicle location information of the multiple transport vehicles and the warehouse location information are respectively input into the arrival predictor to obtain multiple initial vehicle arrival times;

[0054] Multiple driver IDs of multiple transport vehicles are retrieved, and multiple time correction coefficients are obtained based on the multiple driver IDs. The initial arrival times of the multiple vehicles are then corrected to obtain multiple vehicle arrival times.

[0055] First, the system acquires real-time vehicle location information and operational requirements for multiple transport vehicles. Vehicle location information refers to the geographic coordinates uploaded in real-time by the transport vehicles via their onboard GPS devices, including dynamic data such as current latitude and longitude, speed, and direction of travel. Operational requirements information refers to the core requirements data for the current transport task, including the target storage area, cargo type, and estimated loading / unloading volume, used to assist in subsequent scheduling and matching. The system collects the location information of each transport vehicle in real-time through onboard terminals, retrieves the operational requirements information of each vehicle through the task dispatch system, and transmits it synchronously to the dispatch system. For example, real-time acquisition of vehicle location information for three vehicles and retrieval of operational requirements: V1: Current location: 30.1334°N, 120.5678°E, speed 40km / h; Operational requirements: Target storage area A1, 5 tons of refrigerated cargo. V2: Current location: 30.1234°N, 120.5878°E, speed 30km / h; Operational requirements: Target storage area A2, 8 tons of boxed cargo. V3: Current location: 30.1134°N, 120.5678°E, driving speed 35km / h; Operation requirements: target storage area A3, 10 tons of household appliances.

[0056] Secondly, the warehouse location information of the target warehouse is obtained. Warehouse location information refers to the precise geographical coordinates of the loading and unloading area of ​​the target warehouse, which serves as the benchmark for calculating the distance from the vehicle to the warehouse. The registered location information of the target warehouse is retrieved from the warehouse basic information database of the logistics park management system; the coordinates are then calibrated on-site using GPS positioning equipment to finally determine the warehouse location information and store it in the dispatch system. For example, the registered coordinates of warehouse A are 30.1230°N, 120.5675°E. After on-site GPS calibration, the final warehouse location information is determined to be: 30.1234°N, 120.5678°E.

[0057] Furthermore, the vehicle location information of the multiple transport vehicles and the warehouse location information are input into the arrival predictor to obtain multiple initial vehicle arrival times. The arrival predictor is an intelligent prediction module integrating a real-time traffic interface and a distance-time model. It first calculates the shortest travel distance from the vehicle to the warehouse, and then combines real-time traffic conditions and historical travel speed data for the same road segment to output the initial arrival time. The vehicle location information and warehouse location information are input into the arrival predictor. The predictor first calculates the straight-line distance using the Haversine formula, then obtains the road congestion coefficient using the real-time map traffic interface, and estimates the travel time using a long short-term memory network, finally outputting the initial vehicle arrival time. Simultaneously, the operational demand information of each vehicle is recorded to provide data support for subsequent platform matching. For example, inputting the above information into the arrival predictor, combined with real-time traffic conditions, outputs the following initial arrival times: V1 initial arrival time: 12min (current time 9:00, estimated arrival time 9:12); V2 initial arrival time: 14min (estimated arrival time 9:14); V3 initial arrival time: 10min (estimated arrival time 9:10).

[0058] Finally, multiple driver IDs of multiple transport vehicles are retrieved, and multiple time correction coefficients are obtained based on the multiple driver IDs. The arrival times of the multiple initial vehicles are then corrected to obtain the multiple vehicle arrival times.

[0059] First, retrieve the driver IDs for multiple transport vehicles. Each driver ID is a unique identifier assigned to them, linked to their historical driving data, driving habits, and other information. For example, driver ID for vehicle V1 is 001, driver ID for vehicle V2 is 002, and driver ID for vehicle V3 is 003.

[0060] Secondly, multiple time correction coefficients are obtained based on the multiple driver numbers.

[0061] Among them, multiple time correction coefficients are obtained based on the multiple driver IDs, including:

[0062] Extract the first driver ID from the plurality of driver IDs, and obtain the first historical predicted arrival time set and the first historical actual arrival time set based on the first driver ID;

[0063] The ratios of each historical actual arrival time in the first historical actual arrival time set to the corresponding historical predicted arrival time in the first historical predicted arrival time set are calculated to obtain multiple time ratios.

[0064] The average of the multiple time ratios is calculated to obtain the first time correction coefficient corresponding to the first driver number;

[0065] Following the method of obtaining the first time correction coefficient corresponding to the first driver's number, the time correction coefficients corresponding to the other driver's numbers are obtained, resulting in multiple time correction coefficients.

[0066] First, the first driver ID is extracted from the multiple driver IDs. Based on the first driver ID, the first historical predicted arrival time set and the first historical actual arrival time set are obtained. The first driver ID refers to the first calculation object randomly selected from all the driver IDs of the transport vehicles, used to establish the calculation template for the time correction coefficient. Subsequent driver IDs are calculated according to the same logic. The first historical predicted arrival time set refers to the dataset consisting of all historical predicted arrival times of the driver's vehicles to the target warehouse within the past 3 months, associated with the first driver ID. Each data point corresponds to one independent transport task. The first historical actual arrival time set refers to the dataset consisting of the actual arrival times of the driver at the target warehouse, which correspond one-to-one with the first historical predicted arrival time set. Each data point is matched with the record in the predicted time set according to the transport task sequence number. One driver ID is randomly selected from the collected driver IDs as the first driver ID. The driver ID is linked to the driver's historical driving database. The filtering criteria are transportation records within the last 3 months, with the destination of the transportation task being the target warehouse and the task status being completed. The historical predicted arrival time and historical actual arrival time are extracted from each record that meets the criteria, and the first historical predicted arrival time set and the first historical actual arrival time set are formed respectively to ensure that the number of records in the two sets of data is consistent and corresponds one-to-one.

[0067] For example, driver number 001 of V1 is randomly selected as the first driver number. Completed transportation tasks of this driver to warehouse A within the past 3 months are filtered, with a total of 8 valid records. Two sets of time are extracted: the first historical predicted arrival time set: [13,12,14,12,13,11,14,12]min; the first historical actual arrival time set: [12,11,13,11,12,10,13,11]min.

[0068] Next, the ratios of each historical actual arrival time in the first historical actual arrival time set to the corresponding historical predicted arrival time in the first historical predicted arrival time set are calculated to obtain multiple time ratios. The time ratio refers to the ratio of the historical actual arrival time to the corresponding historical predicted arrival time for a single historical transportation task. It reflects the degree of deviation between the driver's actual travel time and the predicted time. A ratio greater than 1 indicates that the actual time is longer than the predicted time, and a ratio less than 1 indicates that the actual time is shorter than the predicted time. Based on the transportation task sequence number, the i-th data point in the first historical actual arrival time set is matched one-to-one with the i-th data point in the first historical predicted arrival time set. Time ratio = historical actual arrival time / historical predicted arrival time. This calculation is performed for each matched data point to obtain multiple time ratios, the same number as the number of historical records.

[0069] For example, time ratios are generated for each of the above 8 matching data: 1st: 12 / 13≈0.92; 2nd: 11 / 12≈0.92; 3rd: 13 / 14≈0.93; 4th: 11 / 12≈0.92; 5th: 12 / 13≈0.92; 6th: 10 / 11≈0.91; 7th: 13 / 14≈0.93; 8th: 11 / 12≈0.92; finally, 8 time ratios are obtained: [0.92, 0.92, 0.93, 0.92, 0.92, 0.91, 0.93, 0.92].

[0070] Then, the average of the multiple time ratios is calculated to obtain the first time correction coefficient corresponding to the first driver's number. The average calculation refers to the process of performing an arithmetic mean operation on multiple time ratios; outliers must be removed before the calculation to ensure the stability and reliability of the coefficient. The first time correction coefficient is a personalized parameter obtained by averaging the time ratios corresponding to the first driver's number, used to correct the initial arrival time of the vehicle driven by that driver. Preliminary statistics are performed on the multiple time ratios, and the average is calculated. and standard deviation ,use The principle is to identify outliers (i.e., values ​​less than 10 ... or greater than The ratios of the remaining effective time ratios are then discarded. The arithmetic mean of the remaining effective time ratios is calculated using the following formula: For example, statistical analysis of the above eight time ratios yields the mean... Standard deviation All ratios are within the range of μ ± 3σ, with no outliers. The arithmetic mean of the effective ratios is calculated as follows: The first time correction coefficient corresponding to the first driver number 001 is determined to be 0.92.

[0071] Furthermore, following the same method as obtaining the first time correction coefficient corresponding to the first driver's number, the time correction coefficients corresponding to the remaining driver's numbers are obtained, resulting in multiple time correction coefficients. The remaining driver's numbers refer to the driver's numbers of all other transport vehicles participating in this dispatch, excluding the first driver's number. The calculation logic for their time correction coefficients is completely consistent with that of the first driver's number. Each of the remaining driver's numbers is selected sequentially from the multiple driver's numbers and used as the current driver's number; the above process is repeated: obtaining the historical predicted / actual arrival time set corresponding to the current driver's number, calculating the time ratio, and averaging to obtain the correction coefficient; the time correction coefficients corresponding to all driver's numbers are collected to form a set of multiple time correction coefficients. For example, for driver number 002, extract 8 valid historical records of this driver's trips to warehouse A within the past 3 months. The historical predicted arrival time set is [15,14,16,14,15,13,16,14], and the historical actual arrival time set is [20,20,22,21,20,22]. Calculate the time ratio: [15,14,16,14,15,13,16,15], with no outliers. Calculate the mean: The second time correction coefficient is determined to be Similarly, the first time correction coefficient corresponding to the third driver's number 003 is 0.97. Finally, we obtain a set of three time correction coefficients: {0.92 (001), 1.01 (002), 0.97 (003)}.

[0072] Finally, the initial vehicle arrival times are corrected to obtain multiple vehicle arrival times. The corrected vehicle arrival time = initial vehicle arrival time × corresponding driver's time correction coefficient. The corrected vehicle arrival time integrates real-time traffic conditions and individual driver driving characteristics, serving as time-series data for subsequent platform scheduling. The initial vehicle arrival time refers to the basic arrival time output by the arrival predictor based on the vehicle's real-time location, warehouse location, and real-time traffic conditions, without considering individual driver differences. A one-to-one correspondence is established between transport vehicles, driver numbers, time correction coefficients, and initial arrival times. Through correlation matching, it is ensured that each vehicle accurately corresponds to its unique time correction coefficient and initial arrival time, avoiding matching errors. The initial time of each transport vehicle is corrected and calculated individually. The corrected arrival times of all vehicles are then aggregated to form the final set of multiple vehicle arrival times.

[0073] For example, establish a complete correspondence between three vehicles: V1: First driver ID 001, first time correction coefficient 0.92, initial arrival time 12 min; V2: Second driver ID 002, second time correction coefficient 1.01, initial arrival time 14 min; V3: Third driver ID 003, third time correction coefficient 0.97, initial arrival time 10 min. Calculate the corrected arrival time for each vehicle: V1 corrected arrival time = 12 × 0.92 = 11.04 11 min; V2 corrected arrival time = 14 × 1.01 = 14.14 14 min; Arrival time after V3 correction = 10 × 0.97 = 9.7 10 min. Finally, we obtained a set of arrival times for multiple vehicles: {V1: 11 min (estimated arrival at 9:11), V2: 14 min (estimated arrival at 9:14), V3: 101 min (estimated arrival at 9:10)}.

[0074] In this embodiment of the invention, by accurately acquiring coordinates, predicting real-time traffic conditions, and correcting personalized coefficients, the prediction error of vehicle arrival time is effectively reduced, effectively solving the problem of insufficient accuracy caused by traditional prediction ignoring individual differences among drivers. The synchronously acquired work demand information provides a demand basis for subsequent platform matching, while the accurate arrival time provides a reliable time series benchmark for subsequent conflict detection, avoiding platform resource idleness or congestion caused by time deviation, and laying a time series foundation for constructing a reasonable vehicle-platform matching scheme.

[0075] S300: Generate an initial vehicle platform matching scheme based on the distance matrix and the operation requirement information, and perform conflict detection on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles to obtain the conflict detection results.

[0076] In this embodiment of the invention, an initial vehicle platform matching scheme is generated based on the distance matrix and the operational demand information. Conflict detection is then performed on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles to obtain the conflict detection results. The rationality of the initial vehicle platform matching and the timeliness of conflict prediction are crucial to avoiding idle or congested platform resources. If only manual experience is relied upon to allocate idle platforms without considering the distance relationship between the vehicle's target storage area and the platform, it can easily lead to excessively long transportation distances after vehicle-platform matching, increasing internal handling costs. Conflict detection relies on real-time manual inspections, which can only detect existing congestion and cannot identify overlapping times when different vehicles occupy the same platform, resulting in a high rate of passive scheduling adjustments. Therefore, it is necessary to construct a distance-demand-oriented initial matching and a time-dimensional early conflict detection process to improve the rationality of matching from the source and avoid potential conflicts.

[0077] Step S300 in the method provided in this embodiment of the invention includes:

[0078] The job requirement information includes the target storage area and job duration;

[0079] Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated, which includes multiple initial vehicle platform pairs.

[0080] Based on the arrival times of the multiple vehicles and the operating time of each transport vehicle, calculate the occupied areas of multiple platforms;

[0081] Conflict detection is performed on the multiple initial vehicle platform pairs based on the multiple platform occupancy intervals to obtain conflict detection results.

[0082] First, the operational requirements information includes the target storage area and the operational duration. Operational requirements information refers to the basis for scheduling the vehicles for this transportation task, containing two key data points: the target storage area and the operational duration. The target storage area refers to the specific area where the goods need to be stored, and the operational duration refers to the estimated time for the vehicles to complete loading and unloading at the platform. For example, retrieve and confirm the operational requirements of three vehicles: V1: Transporting refrigerated fresh produce; fresh produce requires rapid loading and unloading, relatively short time; target storage area is A1; cargo weight is 5 tons; estimated operational duration is 15 minutes; V2: Transporting boxed daily necessities; standardized packaging; moderate loading and unloading efficiency; target storage area is A2; cargo weight is 8 tons; estimated operational duration is 18 minutes; V3: Transporting heavy household appliances; requiring hoisting assistance; relatively long time; target storage area is A3; cargo weight is 10 tons; estimated operational duration is 20 minutes.

[0083] Secondly, based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated, which includes multiple initial vehicle platform pairs.

[0084] Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated. This initial vehicle platform matching scheme includes multiple initial vehicle platform pairs, including:

[0085] Traverse multiple transport vehicles, obtain the first transport vehicle, and acquire the first target storage area of ​​the first transport vehicle;

[0086] Based on the first target storage area, the platform with the shortest transportation distance to the first target storage area is found in the distance matrix and used as the first matching platform for the first transport vehicle.

[0087] The first transport vehicle and the first matching platform are combined to form the first initial vehicle-platform pair.

[0088] Continue to generate corresponding initial vehicle platform pairs for the remaining transport vehicles, resulting in multiple initial vehicle platform pairs. Summarize these pairs to generate an initial vehicle platform matching scheme.

[0089] First, multiple transport vehicles are traversed to obtain the first transport vehicle, and its first target storage area is obtained. The first transport vehicle refers to the first computational object selected in traversal order from the multiple transport vehicles participating in this scheduling, used to establish a demonstration process for vehicle-platform matching. Subsequent vehicles are processed sequentially according to the same logic. The first target storage area refers to the specific cargo storage area where the cargo of the first transport vehicle needs to be stored in this transportation task, and is the core basis for querying the distance matrix and matching the optimal platform. A sequential traversal algorithm is used to extract all transport vehicles from the vehicle task list of the scheduling system to form a vehicle set; the first vehicle in the set is selected in ascending order of vehicle number and determined as the first transport vehicle; the target storage area is extracted from the operation requirement information of this vehicle and determined as the first target storage area. For example, if the vehicle set in this scheduling is {V1, V2, V3}, the first vehicle V1 is selected as the first transport vehicle by traversing in ascending order of number; the target storage area is extracted from the operation requirement information of V1 and determined as A1.

[0090] Secondly, based on the first target storage area, the platform with the shortest transportation distance to the first target storage area is found in the distance matrix and designated as the first matching platform for the first transport vehicle. The first matching platform refers to the platform with the shortest transportation distance to the first target storage area in the distance matrix constructed in S100. Its function is to maximize the reduction of the internal handling distance from the vehicle after unloading to the storage area, thereby improving operational efficiency. The platform-cargo storage area distance matrix constructed in step S100 is loaded to clarify the unique distance correspondence between each platform and each storage area in the matrix. Using the first target storage area as the search condition, the transportation distance data from all platforms to that storage area is extracted from the distance matrix to form a distance data list. The distance data list is sorted from smallest to largest value, and the platform corresponding to the smallest value is selected. If multiple platforms correspond to the same minimum distance, the platform with the lowest usage frequency in the past month is selected first and designated as the first matching platform.

[0091] For example, after loading the distance matrix of warehouse A, the distance data from each platform to A1 is extracted using the first target storage area A1 as the retrieval condition: the distance from platform B1 to A1 is 44m, the distance from platform B2 to A1 is 20m, and the distance from platform B3 to A1 is 68m, forming a distance list [44m, 20m, 68m]. After sorting the list, the minimum distance is 20m, and the corresponding platform is B2. B2 is determined to be the first matching platform of V1.

[0092] Based on this, the first transport vehicle and the first matching platform are combined to form the first initial vehicle-platform pair. The first initial vehicle-platform pair refers to the structured combination formed after establishing a unique association between the first transport vehicle and the first matching platform, presented in the form of vehicle number-platform number, and is the basic building block of the initial matching scheme. A vehicle-platform association file is established, clearly recording the number of the first transport vehicle and the number of the first matching platform; the association is encapsulated using a fixed vehicle-platform format to form the first initial vehicle-platform pair; this vehicle-platform pair is stored in a temporary scheme set to prepare for subsequent aggregation of complete schemes. For example, the number of the first transport vehicle V1 and the number of the first matching platform B2 are recorded, and an association is established; this is encapsulated in the format V1-B2, which is the first initial vehicle-platform pair.

[0093] Then, initial vehicle platform pairs are generated for the remaining transport vehicles, resulting in multiple initial vehicle platform pairs. These are then aggregated to generate an initial vehicle platform matching scheme. The remaining transport vehicles refer to all transport vehicles in the vehicle set that are to be matched with platforms other than the first transport vehicle. Their matching process is completely consistent with that of the first transport vehicle, ensuring the uniformity and fairness of the matching rules. The initial vehicle platform matching scheme is a complete resource allocation scheme formed by integrating and sorting the initial vehicle platform pairs corresponding to all transport vehicles. It is the core input basis for the subsequent conflict detection stage. For example, the first transport vehicle, which has already been matched, is removed from the vehicle set, and the remaining vehicles are selected sequentially as the current transport vehicles. For each current transport vehicle, the above process is repeated: its target storage area is extracted, and the platform with the shortest distance to that storage area is found from the distance matrix, forming the corresponding initial vehicle platform pair and storing it in a temporary scheme set. After all vehicles have been matched, all vehicle platform pairs are extracted from the temporary scheme set, sorted in ascending order by vehicle number, to form the final initial vehicle platform matching scheme.

[0094] For example, processing vehicle V2: extracting the target storage area of ​​V2 as A2, querying the distance from each platform to A2 from the distance matrix: B1 to A2 is 24m, B2 to A2 is 2m, B3 to A2 is 52m; the minimum distance is 2m, corresponding to platform B2, forming the second initial vehicle platform pair V2-B2; processing vehicle V3: extracting the target storage area of ​​V3 as A3, querying the distance from each platform to A3 from the distance matrix: B1 to A3 is 56m, B2 to A3 is 32m, B3 to A3 is 32m; there are two minimum distances (32m) corresponding to platforms B2 and B3, querying the occupancy rate in the past month: B3 occupancy rate 45% < B2 occupancy rate 60%, selecting B3 as the matching platform, forming the third initial vehicle platform pair V3-B3; after sorting by vehicle number in ascending order, the initial vehicle platform matching schemes are summarized as: V1-B2, V2-B2, V3-B3.

[0095] Furthermore, based on the arrival times of the multiple vehicles and the operating time of each transport vehicle, multiple platform occupancy intervals are calculated. A platform occupancy interval refers to the time window during which a single vehicle occupies a particular platform, starting with the vehicle's corrected arrival time and ending with the sum of the arrival time and operating time, accurately reflecting the platform's time occupancy status. The arrival times of each vehicle after S200 correction are retrieved; the start time = corrected arrival time, and the end time = corrected arrival time + operating time. Occupancy intervals are calculated for each vehicle. A table linking platforms, vehicles, and occupancy intervals is established, clearly defining which vehicles occupy each platform and their corresponding time ranges. For example, the current base time is 9:00, and the corrected arrival times are V1: 9:11; V2: 9:14; V3: 9:10. The platform occupancy interval is calculated for each car: V1-B2: start time = 9:11, end time = 9:11 + 15min = 9:26, the occupancy interval is 9:11-9:26; V2-B2: start time = 9:14, end time = 9:14 + 18min = 9:32, the occupancy interval is 9:14-9:32; V3-B3: start time = 9:10, end time = 9:10 + 22min = 9:32, the occupancy interval is 9:10-9:32.

[0096] Finally, conflict detection is performed on the multiple initial vehicle platform pairs based on the multiple platform occupancy intervals to obtain conflict detection results. Conflict detection refers to verifying the time overlap of multiple occupancy intervals of the same platform. If there is an overlap where the previous vehicle has not finished loading / unloading and the subsequent vehicle has already arrived, it is determined to be a conflict. The conflict detection result is a structured report containing whether a conflict exists, the conflicting platform, the conflicting vehicle pair, and the conflict time range, providing a basis for subsequent optimization. Grouping by platform dimension, all vehicle occupancy intervals associated with each platform are extracted; the occupancy intervals of the same platform are sorted by start time, and the previous end time and the next start time of adjacent intervals are verified one by one: if the previous end time > the next start time, it is determined to be a conflict; the conflict information is summarized to generate conflict detection results.

[0097] For example, the occupancy intervals are checked by platform group: Platform B2: The occupancy intervals are V1 (9:11-9:26) and V2 (9:14-9:32). The end time of the previous vehicle V1 is 9:26, which is greater than the start time of the next vehicle V2, which is 9:14, so there is a time overlap. Platform B3: Only V3 occupies the interval (9:10-9:32), and no other vehicles occupy it, so there is no conflict. Conflict information is determined: The conflicting platform is B2, the conflicting vehicle pair is V1-V2, and the conflicting time period is 9:14-9:26. Conflict detection result: The current initial vehicle platform matching scheme of warehouse A has a conflict, specifically: Platform B2 has overlapping occupancy intervals of V1 and V2, the conflicting time period is 9:14-9:26, and the conflicting vehicle pair is V1-V2.

[0098] In this embodiment of the invention, an initial vehicle platform matching scheme is generated by combining a distance matrix with operational requirements, thereby improving the scientific nature of the matching. By constructing platform occupancy intervals based on precise vehicle arrival times and operational durations, platform occupancy conflicts can be predicted in advance, avoiding the passive adjustment situation after a conflict occurs in the traditional model. At the same time, the structured conflict detection results clarify the core elements of the conflict, providing a precise basis for subsequent scheme optimization and effectively ensuring the efficient utilization of platform resources.

[0099] S400: When there are conflicting vehicle platform pairs in the conflict detection results, the initial vehicle platform matching scheme is optimized according to the distance matrix to generate an optimized vehicle platform matching scheme for logistics park platform management.

[0100] In this embodiment of the invention, when conflicting vehicle-platform pairs are detected in the conflict detection results, the initial vehicle-platform matching scheme is optimized based on the distance matrix to generate an optimized vehicle-platform matching scheme for platform management in the logistics park. When platform occupancy conflicts exist in the initial matching scheme, simply adjusting the vehicle-platform allocation can easily lead to increased transportation distances or excessively long vehicle waiting times, affecting overall scheduling efficiency and failing to meet the real-time management needs of the logistics park. Therefore, a full-process optimization mechanism needs to be constructed, encompassing accurate conflict identification, dynamic space delineation, iterative scheme optimization, and real-time instruction issuance, to minimize scheduling costs while avoiding conflicts and ensuring the efficiency and stability of platform management.

[0101] Step S400 in the method provided in this embodiment of the invention includes:

[0102] Extract the conflict vehicle platform pairs from the conflict detection results and identify multiple conflict vehicles that have been in conflict;

[0103] The original start time of the multiple conflicting vehicles is obtained. The earliest original start time is used as the reference point, and an allocable time window is formed by extending it backward according to the dynamic extension duration. The dynamic extension duration is the product of the number of conflicting vehicles and the average operation time of a single vehicle.

[0104] A platform-time two-dimensional space is established based on the allocable time window and multiple platforms, and the platform-time combination that has been occupied by scheduled vehicles is identified as the occupied subspace in the platform-time two-dimensional space.

[0105] In the platform-time two-dimensional space, the occupied subspace is listed as a forbidden subspace to obtain the platform-time allocable space;

[0106] Within the platform-time allocable space, platforms are reallocated for the multiple conflicting vehicles to generate a first candidate solution;

[0107] The transport distance increment of each conflicting vehicle in the first candidate scheme is calculated based on the distance matrix, and the waiting time increment of each conflicting vehicle is calculated to calculate the first fitness.

[0108] Iteratively generate candidate solutions and calculate their fitness until convergence, then select the candidate solution with the highest fitness as the conflict resolution solution;

[0109] The initial vehicle platform matching scheme is optimized based on the conflict resolution to generate an optimized vehicle platform matching scheme.

[0110] Based on the optimized vehicle platform matching scheme, dispatch instructions for each transport vehicle are generated, and the dispatch instructions for each transport vehicle are sent to each transport vehicle through the edge collaborative network to execute the platform management of the logistics park.

[0111] First, conflicting vehicle platform pairs are extracted from the conflict detection results to identify multiple conflicting vehicles. A conflicting vehicle platform pair refers to a vehicle-platform combination with overlapping occupancy in the conflict detection results; it is the core processing object for conflict resolution. Conflicting vehicles refer to all transport vehicles involved in the conflict, i.e., all vehicles involved in the conflicting vehicle platform pair, which need to be centrally redistributed. From the conflict detection results of S300, the set of vehicle platform pairs with occupancy conflicts is extracted; all vehicles involved in the conflict are then summarized and marked as conflicting vehicles. For example, from the conflict detection results, the conflicting vehicle platform pair {V1-B2, V2-B2} is extracted, and the conflicting vehicles are identified as V1 and V2.

[0112] Secondly, the original start times of the multiple conflicting vehicles are obtained. Using the earliest original start time as the baseline, an allocable time window is formed by extending the window forward based on a dynamic extension duration. The dynamic extension duration is the product of the number of conflicting vehicles and the average operation time per vehicle. The original start time refers to the starting time of the conflicting vehicles' platform occupancy in the initial plan, i.e., the corrected vehicle arrival time. The baseline is the earliest time among the original start times of the conflicting vehicles, serving as the starting node of the allocable time window. The dynamic extension duration is the length of time calculated based on the number of conflicting vehicles and the average operation time per vehicle, used to ensure adjustment space for conflict resolution. The allocable time window is a continuous time period formed by extending the dynamic extension duration forward from the baseline, representing the time range for subsequent platform allocation. The original start times of the conflicting vehicles in the initial plan are retrieved; the earliest time is selected as the baseline; the dynamic extension duration = number of conflicting vehicles × average operation time per vehicle, extending backward from the baseline to form an allocable time window.

[0113] For example, retrieve the original start time of the conflicting vehicles: V1 is 9:11 and V2 is 9:14. Use the earliest original start time of 9:11 as the reference point; there are 2 conflicting vehicles, the average operation time per vehicle is 16.5 min, and the dynamic extension time = 2 × 16.5 min = 33 min; the allocable time window is 33 min after the reference point 9:11, that is, 9:11-9:44.

[0114] Next, a platform-time two-dimensional space is established based on the allocable time window and multiple platforms. Within this space, platform-time combinations already occupied by scheduled vehicles are identified as occupied subspaces. The platform-time two-dimensional space is a two-dimensional scheduling space constructed with available platforms as the horizontal axis and allocable time windows as the vertical axis, visually representing the combination relationship between platforms and time. An occupied subspace refers to the area within the platform-time two-dimensional space that has been occupied by non-conflicting vehicles; this space cannot be reassigned. The horizontal axis represents all available platforms in the target warehouse, and the vertical axis represents the allocable time window, constructing the platform-time two-dimensional space. The platform occupancy intervals of non-conflicting vehicles are retrieved, and the corresponding platform-time combinations are marked in the two-dimensional space, which are the occupied subspaces. For example, construct a platform-time two-dimensional space for warehouse A: the horizontal axis is B1, B2, B3, and the vertical axis is 9:11-9:44; the occupancy range of non-conflicting vehicle V3 is 9:10-9:32, and the corresponding combination in the two-dimensional space is B3-9:11-9:32. Mark this area as the occupied subspace.

[0115] Furthermore, in the platform-time two-dimensional space, the occupied subspaces are listed as forbidden subspaces, resulting in platform-time allocable space. Forbidden subspaces are areas where occupied subspaces are explicitly designated as unallocable, preventing conflicting vehicles from clashing with non-conflicting vehicles. Platform-time allocable space refers to the remaining free area in the platform-time two-dimensional space after removing forbidden subspaces; it is the resource pool for the redistribution of conflicting vehicles. In the platform-time two-dimensional space, all identified occupied subspaces are listed as forbidden subspaces; after removing forbidden subspaces, the remaining platform-time combination areas in the two-dimensional space constitute the platform-time allocable space. For example, B3-9:11-9:32 is listed as a forbidden subspace; after removal, the allocable space includes: B1 (9:11-9:44), B2 (9:11-9:44), and B3 (9:33-9:44).

[0116] Then, within the platform-time allocable space, platforms are reassigned for the multiple conflicting vehicles to generate a first candidate solution. The first candidate solution refers to the first conflict resolution candidate solution formed after rematching platform-time combinations for all conflicting vehicles within the platform-time allocable space. Based on the allocable space and considering the target storage area requirements of the conflicting vehicles, non-overlapping platforms and time segments are randomly assigned to each conflicting vehicle; ensuring that all assigned platform-time combinations are within the allocable space and that there is no time overlap between conflicting vehicles, thus forming the first candidate solution. For example, within the allocable space, V1 and V2 are reassigned: V1 is assigned to B1, time segment 9:11-9:26 (operation duration 15min); V2 remains in B2, time segment 9:27-9:45 (operation duration 18min, 9:27-9:44 within the allocable window); first candidate solution: {V1-B1 (9:11-9:26), V2-B2 (9:27-9:45)}.

[0117] Based on this, the transport distance increment of each conflicting vehicle in the first candidate scheme is calculated according to the distance matrix, and the waiting time increment of each conflicting vehicle is calculated to calculate the first fitness.

[0118] Specifically, the calculation of the transport distance increment for each conflicting vehicle in the first candidate scheme based on the distance matrix, the calculation of the waiting time increment for each conflicting vehicle, and the calculation of the first fitness include:

[0119] Extract the platforms for the reassignment of each conflicting vehicle from the first candidate scheme, combine them with the target storage area of ​​each conflicting vehicle, and find the corresponding transportation distance value in the distance matrix to obtain the transportation distance value for the reassignment of each conflicting vehicle.

[0120] Calculate the difference between the reallocated transport distance value and the original transport distance value for each conflicting vehicle to obtain the transport distance increment for each conflicting vehicle, and sum them to obtain the total transport distance increment;

[0121] Calculate the difference between the reassigned start time of each conflicting vehicle and the original start time of the operation to obtain the waiting time increment of each conflicting vehicle, and sum them to obtain the total waiting time increment;

[0122] The increments of total transport distance and total waiting time are normalized to obtain normalized distance increments and normalized time increments.

[0123] The normalized distance increment and normalized time increment are weighted and summed according to the preset weight coefficients, and the reciprocal of the weighted sum is performed to obtain the first fitness.

[0124] First, the platforms reassigned to each conflicting vehicle are extracted from the first candidate scheme. Combined with the target storage area of ​​each conflicting vehicle, the corresponding transport distance value is searched in the distance matrix to obtain the reassigned transport distance value for each conflicting vehicle. The reassigned transport distance value refers to the transport distance from the re-matched platform to the target storage area of ​​the conflicting vehicle in the first candidate scheme, obtained by querying the distance matrix constructed in S100. The originally assigned transport distance value refers to the transport distance from the original matched platform to the target storage area of ​​the conflicting vehicle in the initial matching scheme, serving as the baseline data for calculating the distance increment. The platform number reassigned to each conflicting vehicle is extracted from the first candidate scheme; the target storage area of ​​each conflicting vehicle is associated, and the corresponding transport distance value is searched in the distance matrix using the reassigned platform-target storage area as the search condition; the originally assigned transport distance value of each conflicting vehicle in the initial scheme is retrieved simultaneously for later calculation.

[0125] For example, extract the reassignment platforms for conflicting vehicles in the first candidate scheme: V1 is reassigned to B1, and V2 remains in B2; query the distance matrix to obtain the reassigned transportation distance values: V1's target storage area is A1, and the distance from B1 to A1 is 44m; V2's target storage area is A2, and the distance from B2 to A2 is 2m; retrieve the original assigned transportation distance values: V1's original platform B2 to A1 is 20m, and V2's original platform B2 to A2 is 2m.

[0126] Secondly, the difference between the redistributed transport distance value and the original transport distance value for each conflicting vehicle is calculated to obtain the transport distance increment for each conflicting vehicle. These increments are then summed to obtain the total transport distance increment. The transport distance increment refers to the difference between the redistributed transport distance and the original transport distance for a single conflicting vehicle; a positive number indicates an increase in distance, while zero indicates no change. The total transport distance increment is the sum of the transport distance increments for all conflicting vehicles, reflecting the overall change in distance cost of the candidate solutions. Transport distance increment = redistributed transport distance value - original transport distance value. The distance increment for each conflicting vehicle is calculated; the total transport distance increment is obtained by summing the transport distance increments for all conflicting vehicles. For example, calculating the transport distance increment for each vehicle: V1 increment = 44m - 20m = 24m; V2 increment = 2m - 2m = 0m; calculating the total transport distance increment: 24m + 0m = 24m.

[0127] Subsequently, the difference between the reassigned start time of each conflicting vehicle and its original start time is calculated to obtain the waiting time increment for each conflicting vehicle. These increments are then summed to obtain the total waiting time increment. The waiting time increment refers to the difference between the reassigned start time of a single conflicting vehicle and its original start time; a positive number indicates that waiting is required, and zero indicates that no waiting is required. The total waiting time increment is the sum of the waiting time increments of all conflicting vehicles, reflecting the overall time cost change of the candidate schemes. The reassigned start times of each conflicting vehicle are extracted from the first candidate scheme, and the original start times are retrieved simultaneously. The waiting time increment = reassigned start time - original start time. The time increment for each conflicting vehicle is calculated; the total waiting time increment is obtained by summing the waiting time increments of all conflicting vehicles. For example, extract the start time: V1 original start time 9:11, reallocated start time 9:11; V2 original start time 9:14, reallocated start time 9:27; calculate the waiting time increment for each vehicle: V1 increment = 9:11 - 9:11 = 0 min; V2 increment = 9:27 - 9:14 = 13 min; calculate the total waiting time increment: 0 min + 13 min = 13 min.

[0128] Furthermore, the total transportation distance increment and total waiting time increment are normalized to obtain normalized distance increment and normalized time increment. Normalization involves converting the total transportation distance increment and total waiting time increment into dimensionless values ​​within the range [0,1], eliminating the dimensional differences between the two and facilitating subsequent weighted calculations. Normalized distance increment / normalized time increment refers to the normalized total distance increment and total time increment; the closer the value is to 0, the lower the corresponding cost. The maximum transportation distance increment threshold and maximum waiting time increment threshold preset by the logistics park are retrieved; the normalized value = actual total increment / corresponding maximum threshold, and the normalized distance increment and normalized time increment are calculated respectively. For example, the park's preset thresholds are retrieved: maximum transportation distance increment threshold 50m, maximum waiting time increment threshold 30min; normalized distance increment = 24m / 50m = 0.48; normalized time increment = 13min / 30min ≈ 0.433.

[0129] Finally, the normalized distance increment and normalized time increment are weighted and summed according to preset weight coefficients, and the reciprocal of the weighted sum is taken to obtain the first fitness. The preset weight coefficients refer to the distance and time weights set according to the park's scheduling priority, used to balance the importance of both in the scheme evaluation, with a total weight sum of 1. The smaller the weighted sum, the lower the overall cost of the candidate scheme. The first fitness, the reciprocal of the weighted sum, is the core indicator for evaluating the merits of candidate schemes; the larger the value, the better the overall performance of the scheme. The preset weight coefficients are retrieved; the weighted sum = normalized distance increment × distance weight + normalized time increment × time weight, calculating the overall cost; the reciprocal of the weighted sum is taken to obtain the first fitness. For example, setting a distance weight of 0.5 and a time weight of 0.5 to adapt to the scheduling requirements of balancing distance and time in the park; substituting the weights to calculate the weighted sum result: 0.48×0.5+0.433×0.5=0.4565; first fitness = 1 / 0.4565≈2.19, the larger the value, the smaller the overall cost of the solution and the higher the fitness.

[0130] Subsequently, candidate solutions are iteratively generated and their fitness is calculated until convergence. The candidate solution with the highest fitness is selected as the conflict resolution solution. Iteration refers to repeatedly executing the process of generating candidate solutions and calculating fitness, continuously exploring better solutions, and reducing the overall scheduling cost through multiple attempts. The convergence threshold is a preset critical value for determining when iteration stops. When the difference in optimal fitness between two adjacent iterations is less than this threshold, it indicates that the solution is close to optimal, and iteration stops. The conflict resolution solution is the candidate solution with the highest fitness after iteration convergence, which can avoid conflicts while minimizing the overall cost of the increase in total transportation distance and the increase in total waiting time. Based on the park's scheduling accuracy requirements, a convergence threshold is set, and an iteration count counter and optimal fitness cache are initialized. The complete process of generating candidate solutions and calculating fitness is repeated. Each time, the platform allocation and operation start time of conflicting vehicles are randomly adjusted to generate new candidate solutions and calculate the corresponding fitness. The fitness generated each time is compared with the optimal fitness in the cache, and the cache is updated to the current maximum fitness. The difference between the optimal fitness of two adjacent iterations is calculated. If the difference is less than the convergence threshold, the iteration stops. If it is not satisfied, the iteration continues. After the iteration converges, the candidate solution corresponding to the optimal fitness in the cache is determined as the conflict solution.

[0131] For example, the preset convergence threshold is 0.05, the initial optimal fitness cache is 2.19 (the first candidate solution), and the iteration counter starts from 1; the second iteration generates the second candidate solution: V1 is assigned to B2 (9:27-9:42), V2 is assigned to B1 (9:11-9:29), the calculated fitness is 2.12, which does not exceed the current best, and the optimal cache remains 2.19; the third iteration generates the third candidate solution: V1 is assigned to B1 (9:11-9:26), V2 is assigned to B2 (9:27-9:45). The fitness is calculated to be 2.25, and the optimal cache is updated to 2.25. In the fourth iteration, the fourth candidate solution is generated: V1 is assigned to B1 (9:15-9:30), and V2 is assigned to B2 (9:27-9:45). The fitness is calculated to be 2.24. The difference between the two adjacent optimal fitnesss is 2.25-2.24=0.01, which is less than the convergence threshold of 0.05, so the iteration stops. The third candidate solution with the highest fitness is determined as the conflict solution: {V1-B1 (9:11-9:26), V2-B2 (9:27-9:45)}.

[0132] Furthermore, the initial vehicle platform matching scheme is optimized based on the conflict resolution method to generate an optimized vehicle platform matching scheme. The optimized vehicle platform matching scheme integrates the conflict resolution method into the initial vehicle platform matching scheme, replacing the original conflicting vehicle platform pairs to form a complete, conflict-free resource allocation scheme, which is the core basis for subsequent scheduling execution. Conflicting vehicle platform pair replacement refers to using the new matching relationships of conflicting vehicles from the conflict resolution method to cover the corresponding conflicting combinations in the initial scheme, while keeping the matching relationships of non-conflicting vehicles unchanged. The complete set of initial vehicle platform matching schemes is extracted, and conflicting vehicle platform pairs are selected. The new vehicle-platform combinations from the conflict resolution method are used to replace the conflicting vehicle platform pairs in the initial scheme one by one. The original matching relationships of non-conflicting vehicles in the initial scheme are retained, and all replaced vehicle platform pairs are integrated and sorted to generate the optimized vehicle platform matching scheme.

[0133] For example, the initial vehicle platform matching scheme is {V1-B2, V2-B2, V3-B3}, where the conflicting vehicle platform pairs are {V1-B2, V2-B2}. The above conflicting pairs are replaced with {V1-B1 (9:11-9:26), V2-B2 (9:27-9:45)} from the conflict solution. The original matching relationship of the non-conflicting vehicle V3, {V3-B3 (9:10-9:32)}, is retained. After integration, the optimized vehicle platform matching scheme is obtained: {V1-B1 (9:11-9:26), V2-B2 (9:27-9:45), V3-B3 (9:10-9:32)}.

[0134] Finally, based on the optimized vehicle platform matching scheme, dispatch instructions for each transport vehicle are generated. These instructions are then distributed to each transport vehicle via an edge collaborative network to execute platform control within the logistics park. The dispatch instructions are structured execution instructions generated based on the optimized vehicle platform matching scheme, containing core information such as vehicle number, designated platform, start time, and duration, ensuring clear guidelines for vehicle operations. The edge collaborative network is a low-latency, high-reliability communication network deployed within the logistics park, enabling real-time transmission of dispatch instructions and avoiding dispatch chaos caused by instruction delays. For each transport vehicle in the optimized scheme, information such as vehicle number, designated platform, start time, and duration is extracted to generate standardized dispatch instructions. These instructions are then distributed to the corresponding vehicle's onboard terminal via the edge collaborative network, ensuring real-time and accurate transmission. Upon receiving the instructions, the onboard terminal informs the driver via pop-up windows and voice prompts, guiding the vehicle to the designated platform. The park's dispatch center monitors the operation progress of each vehicle in real time, ensuring that the vehicles complete their operations within the specified time, and implementing platform control throughout the process.

[0135] For example, standardized dispatch instructions are generated as follows: V1: Vehicle number V1, designated platform B1, operation start time 9:11, operation duration 15 minutes, operation end time 9:26; V2: Vehicle number V2, designated platform B2, operation start time 9:27, operation duration 18 minutes, operation end time 9:45; V3: Vehicle number V3, designated platform B3, operation start time 9:10, operation duration 22 minutes, operation end time 9:32. The above instructions are sent to the on-board terminals of V1, V2, and V3 respectively through the edge collaborative network. After receiving the instructions, the driver drives the vehicle to the corresponding platform and completes the loading and unloading operation within the specified time. The park dispatch center monitors the operation process in real time, realizing conflict-free and efficient platform management.

[0136] Specifically, when there are no conflicting vehicle platform pairs in the conflict detection results, the initial vehicle platform matching scheme is used as the optimized vehicle platform matching scheme.

[0137] Furthermore, when no conflicting vehicle platform pairs are found in the conflict detection results, the initial vehicle platform matching scheme is used as the optimized vehicle platform matching scheme. A conflict-free vehicle platform pair refers to a situation where, in the conflict detection results, the platform occupancy intervals corresponding to all initial vehicle platform pairs do not overlap in time, and there is no situation where the same platform is occupied by multiple vehicles simultaneously. When the initial scheme already meets the conflict-free requirement, no iterative optimization is needed; the initial vehicle platform matching scheme is directly identified as the optimized vehicle platform matching scheme, reducing redundant calculations and improving scheduling efficiency. The system receives the conflict detection results output by S300 and determines whether conflicting vehicle platform pairs exist in the results. If the results clearly indicate no conflict, the above conflict optimization process is skipped; the initial vehicle platform matching scheme is directly marked as the final optimized vehicle platform matching scheme, and the process proceeds to the subsequent scheduling instruction generation stage.

[0138] For example, the initial vehicle platform matching scheme is {V1-B2, V2-B1, V3-B3}. The vehicle arrival time and operation duration after S200 correction are retrieved, and the platform occupancy intervals are calculated: V1 (V1-B2): Arrival time 9:11, operation duration 15 minutes, occupancy interval 9:11-9:26; V2 (V2-B1): Arrival time 9:14, operation duration 18 minutes, occupancy interval 9:14-9:32; V3 (V3-B3): Arrival time 9:10, operation duration 22 minutes, occupancy interval 9:10-9:32. Conflict detection results show that platforms B2, B1, and B3 each correspond to only one vehicle, and there is no time overlap between the occupancy intervals. The conflict detection result is a conflict-free vehicle platform pair. No subsequent conflict optimization process is required; the initial scheme {V1-B2, V2-B1, V3-B3} is directly determined as the optimized vehicle platform matching scheme, and subsequent scheduling instructions can be generated according to this scheme.

[0139] In this embodiment of the invention, by accurately extracting conflicting vehicles and dynamically defining allocable time windows, combined with marking taboo subspaces in a platform-time two-dimensional space, a scientific adjustment basis for conflict resolution is provided, avoiding new conflicts with non-conflicting vehicles. Through iterative optimization and fitness evaluation mechanisms, the cost of transportation distance and waiting time is balanced, selecting the conflict solution with the best overall performance, avoiding the subjective defects of traditional manual adjustments. In conflict-free scenarios, the initial solution is directly used, reducing redundant calculations and improving efficiency. Real-time dispatch instructions are issued through an edge collaborative network to ensure accurate implementation of the solution. Overall, efficient conflict resolution and optimized allocation of platform resources are achieved, improving the stability and efficiency of platform management in logistics parks.

[0140] Through the specific implementation methods described above, the embodiments of the present invention achieve the following technical effects:

[0141] This invention provides a method and system for intelligent management and control of logistics park platforms based on edge collaboration. By constructing a precise platform-storage area distance matrix, it provides reliable spatial data support for vehicle platform matching; by combining driver-specific characteristics to correct vehicle arrival times, it improves the accuracy of time-series data; based on distance and operational requirements, it generates an initial matching scheme and simultaneously predicts platform occupancy conflicts, avoiding the passive adjustments of traditional scheduling; in conflict scenarios, it uses two-dimensional spatial analysis and iterative optimization to select the optimal solution; in non-conflict scenarios, it directly uses the initial scheme to reduce redundant calculations; and finally, it issues scheduling instructions in real time through an edge collaborative network. The overall process achieves scientific vehicle platform matching, proactive conflict resolution, optimized resource allocation, and efficient management and control, comprehensively improving the stability and efficiency of logistics park platform management and control.

[0142] Example 2, as Figure 2 As shown, this invention provides an intelligent management and control system for logistics park platforms based on edge collaboration, the system comprising:

[0143] The distance matrix establishment module 11 is used to determine the target warehouse, which has multiple cargo storage areas and multiple platforms, and to establish a distance matrix between the multiple cargo storage areas and multiple platforms.

[0144] Arrival time prediction module 12 is used to obtain vehicle location information and operation demand information of multiple transport vehicles in real time, and predict the arrival time of the multiple vehicles based on the vehicle location information of the multiple transport vehicles.

[0145] The matching conflict detection module 13 is used to generate an initial vehicle platform matching scheme based on the distance matrix and the operation requirement information, and to perform conflict detection on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles, and obtain the conflict detection result.

[0146] The matching scheme optimization module 14 is used to optimize the initial vehicle platform matching scheme according to the distance matrix when there are conflicting vehicle platform pairs in the conflict detection results, generate an optimized vehicle platform matching scheme, and perform platform management in the logistics park.

[0147] In one embodiment, the distance matrix establishment module 11 is further configured to:

[0148] This involves establishing a distance matrix for multiple cargo storage areas and multiple platforms, including:

[0149] Obtain a three-dimensional model of the target warehouse, determine the coordinate range of multiple cargo storage areas and the coordinate positions of multiple platforms in the three-dimensional model, and obtain multiple storage coordinate ranges and multiple platform coordinate positions;

[0150] The first platform is determined from the plurality of platforms, and its coordinate position is obtained;

[0151] Based on the channel layout in the three-dimensional model, calculate the shortest transportation path from the coordinate position of the first platform to each storage coordinate range, and obtain multiple transportation distance values ​​from the first platform to each cargo storage area.

[0152] By obtaining multiple transport distance values ​​from the first platform to each cargo storage area, multiple transport distance values ​​from the remaining platforms to each cargo storage area are obtained, resulting in multiple transport distance values ​​from the platforms to each cargo storage area.

[0153] The transportation distance values ​​from multiple platforms to each cargo storage area are aggregated to construct the distance matrix.

[0154] In one embodiment, the arrival time prediction module 12 is further configured to:

[0155] The method of predicting and obtaining the arrival times of the multiple transport vehicles based on their location information includes:

[0156] Obtain the warehouse location information of the target warehouse;

[0157] The vehicle location information of the multiple transport vehicles and the warehouse location information are respectively input into the arrival predictor to obtain multiple initial vehicle arrival times;

[0158] Multiple driver IDs of multiple transport vehicles are retrieved, and multiple time correction coefficients are obtained based on the multiple driver IDs. The initial arrival times of the multiple vehicles are then corrected to obtain multiple vehicle arrival times.

[0159] Among them, multiple time correction coefficients are obtained based on the multiple driver IDs, including:

[0160] Extract the first driver ID from the plurality of driver IDs, and obtain the first historical predicted arrival time set and the first historical actual arrival time set based on the first driver ID;

[0161] The ratios of each historical actual arrival time in the first historical actual arrival time set to the corresponding historical predicted arrival time in the first historical predicted arrival time set are calculated to obtain multiple time ratios.

[0162] The average of the multiple time ratios is calculated to obtain the first time correction coefficient corresponding to the first driver number;

[0163] Following the method of obtaining the first time correction coefficient corresponding to the first driver's number, the time correction coefficients corresponding to the other driver's numbers are obtained, resulting in multiple time correction coefficients.

[0164] In one embodiment, the matching conflict detection module 13 is further configured to:

[0165] The job requirement information includes the target storage area and job duration;

[0166] Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated, which includes multiple initial vehicle platform pairs.

[0167] Based on the arrival times of the multiple vehicles and the operating time of each transport vehicle, calculate the occupied areas of multiple platforms;

[0168] Conflict detection is performed on the multiple initial vehicle platform pairs based on the multiple platform occupancy intervals to obtain conflict detection results.

[0169] Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated. This initial vehicle platform matching scheme includes multiple initial vehicle platform pairs, including:

[0170] Traverse multiple transport vehicles, obtain the first transport vehicle, and acquire the first target storage area of ​​the first transport vehicle;

[0171] Based on the first target storage area, the platform with the shortest transportation distance to the first target storage area is found in the distance matrix and used as the first matching platform for the first transport vehicle.

[0172] The first transport vehicle and the first matching platform are combined to form the first initial vehicle-platform pair.

[0173] Continue to generate corresponding initial vehicle platform pairs for the remaining transport vehicles, resulting in multiple initial vehicle platform pairs. Summarize these pairs to generate an initial vehicle platform matching scheme.

[0174] In one embodiment, the matching scheme optimization module 14 is further configured to:

[0175] Extract the conflict vehicle platform pairs from the conflict detection results and identify multiple conflict vehicles that have been in conflict;

[0176] The original start time of the multiple conflicting vehicles is obtained. The earliest original start time is used as the reference point, and an allocable time window is formed by extending it backward according to the dynamic extension duration. The dynamic extension duration is the product of the number of conflicting vehicles and the average operation time of a single vehicle.

[0177] A platform-time two-dimensional space is established based on the allocable time window and multiple platforms, and the platform-time combination that has been occupied by scheduled vehicles is identified as the occupied subspace in the platform-time two-dimensional space.

[0178] In the platform-time two-dimensional space, the occupied subspace is listed as a forbidden subspace to obtain the platform-time allocable space;

[0179] Within the platform-time allocable space, platforms are reallocated for the multiple conflicting vehicles to generate a first candidate solution;

[0180] The transport distance increment of each conflicting vehicle in the first candidate scheme is calculated based on the distance matrix, and the waiting time increment of each conflicting vehicle is calculated to calculate the first fitness.

[0181] Iteratively generate candidate solutions and calculate their fitness until convergence, then select the candidate solution with the highest fitness as the conflict resolution solution;

[0182] The initial vehicle platform matching scheme is optimized based on the conflict resolution to generate an optimized vehicle platform matching scheme.

[0183] Based on the optimized vehicle platform matching scheme, dispatch instructions for each transport vehicle are generated, and the dispatch instructions for each transport vehicle are sent to each transport vehicle through the edge collaborative network to execute the platform management of the logistics park.

[0184] Specifically, the calculation of the transport distance increment for each conflicting vehicle in the first candidate scheme based on the distance matrix, the calculation of the waiting time increment for each conflicting vehicle, and the calculation of the first fitness include:

[0185] Extract the platforms for the reassignment of each conflicting vehicle from the first candidate scheme, combine them with the target storage area of ​​each conflicting vehicle, and find the corresponding transportation distance value in the distance matrix to obtain the transportation distance value for the reassignment of each conflicting vehicle.

[0186] Calculate the difference between the reallocated transport distance value and the original transport distance value for each conflicting vehicle to obtain the transport distance increment for each conflicting vehicle, and sum them to obtain the total transport distance increment;

[0187] Calculate the difference between the reassigned start time of each conflicting vehicle and the original start time of the operation to obtain the waiting time increment of each conflicting vehicle, and sum them to obtain the total waiting time increment;

[0188] The increments of total transport distance and total waiting time are normalized to obtain normalized distance increments and normalized time increments.

[0189] The normalized distance increment and normalized time increment are weighted and summed according to the preset weight coefficients, and the reciprocal of the weighted sum is performed to obtain the first fitness.

[0190] Specifically, when there are no conflicting vehicle platform pairs in the conflict detection results, the initial vehicle platform matching scheme is used as the optimized vehicle platform matching scheme.

[0191] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0192] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0193] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.

Claims

1. A method for intelligent management and control of logistics park platforms based on edge collaboration, characterized in that, The method includes: Identify the target warehouse, which has multiple cargo storage areas and multiple platforms, and establish a distance matrix between the multiple cargo storage areas and multiple platforms; The system can acquire the real-time location information and operational requirements of multiple transport vehicles, and predict the arrival time of multiple vehicles based on their location information. An initial vehicle platform matching scheme is generated based on the distance matrix and the operation requirement information, and conflict detection is performed on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles to obtain the conflict detection results. When conflicting vehicle-platform pairs are detected in the conflict detection results, the initial vehicle-platform matching scheme is optimized based on the distance matrix to generate an optimized vehicle-platform matching scheme for platform management in the logistics park, including: Extract the conflict vehicle platform pairs from the conflict detection results and identify multiple conflict vehicles that have been in conflict; The original start time of the multiple conflicting vehicles is obtained. The earliest original start time is used as the reference point, and an allocable time window is formed by extending it backward according to the dynamic extension duration. The dynamic extension duration is the product of the number of conflicting vehicles and the average operation time of a single vehicle. A platform-time two-dimensional space is established based on the allocable time window and multiple platforms, and the platform-time combination that has been occupied by scheduled vehicles is identified as the occupied subspace in the platform-time two-dimensional space. In the platform-time two-dimensional space, the occupied subspace is listed as a forbidden subspace to obtain the platform-time allocable space; Within the platform-time allocable space, platforms are reallocated for the multiple conflicting vehicles to generate a first candidate solution; Calculate the transport distance increment for each conflicting vehicle in the first candidate scheme based on the distance matrix, calculate the waiting time increment for each conflicting vehicle, and calculate the first fitness, including: Extract the platforms for the reassignment of each conflicting vehicle from the first candidate scheme, combine them with the target storage area of ​​each conflicting vehicle, and find the corresponding transportation distance value in the distance matrix to obtain the transportation distance value for the reassignment of each conflicting vehicle. Calculate the difference between the reallocated transport distance value and the original transport distance value for each conflicting vehicle to obtain the transport distance increment for each conflicting vehicle, and sum them to obtain the total transport distance increment; Calculate the difference between the reassigned start time of each conflicting vehicle and the original start time of the operation to obtain the waiting time increment of each conflicting vehicle, and sum them to obtain the total waiting time increment; The increments of total transport distance and total waiting time are normalized to obtain normalized distance increments and normalized time increments. The normalized distance increment and normalized time increment are weighted and summed according to the preset weight coefficients, and the reciprocal of the weighted sum is performed to obtain the first fitness. Iteratively generate candidate solutions and calculate their fitness until convergence, then select the candidate solution with the highest fitness as the conflict resolution solution; The initial vehicle platform matching scheme is optimized based on the conflict resolution to generate an optimized vehicle platform matching scheme. Based on the optimized vehicle platform matching scheme, dispatch instructions for each transport vehicle are generated, and the dispatch instructions for each transport vehicle are sent to each transport vehicle through the edge collaborative network to execute the platform management of the logistics park.

2. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 1, characterized in that, Establish a distance matrix for multiple cargo storage areas and multiple platforms, including: Obtain a three-dimensional model of the target warehouse, determine the coordinate range of multiple cargo storage areas and the coordinate positions of multiple platforms in the three-dimensional model, and obtain multiple storage coordinate ranges and multiple platform coordinate positions; The first platform is determined from the plurality of platforms, and its coordinate position is obtained; Based on the channel layout in the three-dimensional model, calculate the shortest transportation path from the coordinate position of the first platform to each storage coordinate range, and obtain multiple transportation distance values ​​from the first platform to each cargo storage area. By obtaining multiple transport distance values ​​from the first platform to each cargo storage area, multiple transport distance values ​​from the remaining platforms to each cargo storage area are obtained, resulting in multiple transport distance values ​​from the platforms to each cargo storage area. The distance values ​​from multiple platforms to each cargo storage area are aggregated to construct the distance matrix.

3. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 1, characterized in that, The arrival times of the multiple transport vehicles are predicted based on their location information, including: Obtain the warehouse location information of the target warehouse; The vehicle location information of the multiple transport vehicles and the warehouse location information are respectively input into the arrival predictor to obtain multiple initial vehicle arrival times; Multiple driver IDs of multiple transport vehicles are retrieved, and multiple time correction coefficients are obtained based on the multiple driver IDs. The initial arrival times of the multiple vehicles are then corrected to obtain multiple vehicle arrival times.

4. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 3, characterized in that, Multiple time correction coefficients are obtained based on the multiple driver IDs, including: Extract the first driver ID from the plurality of driver IDs, and obtain the first historical predicted arrival time set and the first historical actual arrival time set based on the first driver ID; The ratios of each historical actual arrival time in the first historical actual arrival time set to the corresponding historical predicted arrival time in the first historical predicted arrival time set are calculated to obtain multiple time ratios. The average of the multiple time ratios is calculated to obtain the first time correction coefficient corresponding to the first driver number; Following the method of obtaining the first time correction coefficient corresponding to the first driver's number, the time correction coefficients corresponding to the other driver's numbers are obtained, resulting in multiple time correction coefficients.

5. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 1, characterized in that, An initial vehicle platform matching scheme is generated based on the distance matrix and the operational requirement information. Conflict detection is then performed on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles, and the conflict detection results are obtained, including: The job requirement information includes the target storage area and job duration; Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated, which includes multiple initial vehicle platform pairs. Based on the arrival times of the multiple vehicles and the operating time of each transport vehicle, calculate the occupied areas of multiple platforms; Conflict detection is performed on the multiple initial vehicle platform pairs based on the multiple platform occupancy intervals to obtain conflict detection results.

6. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 5, characterized in that, Based on the distance matrix and the target storage area of ​​each transport vehicle, an initial vehicle platform matching scheme is generated. This initial vehicle platform matching scheme includes multiple initial vehicle platform pairs, including: Traverse multiple transport vehicles, obtain the first transport vehicle, and acquire the first target storage area of ​​the first transport vehicle; Based on the first target storage area, the platform with the shortest transportation distance to the first target storage area is found in the distance matrix and used as the first matching platform for the first transport vehicle. The first transport vehicle and the first matching platform are combined to form the first initial vehicle-platform pair. Continue to generate corresponding initial vehicle platform pairs for the remaining transport vehicles, resulting in multiple initial vehicle platform pairs. Summarize these pairs to generate an initial vehicle platform matching scheme.

7. The intelligent management and control method for logistics park platforms based on edge collaboration according to claim 1, characterized in that, When no conflicting vehicle platform pairs are found in the conflict detection results, the initial vehicle platform matching scheme is used as the optimized vehicle platform matching scheme.

8. An intelligent control system for logistics park platforms based on edge collaboration, characterized in that: The system is used to implement the edge-collaboration-based intelligent management and control method for logistics park platforms according to any one of claims 1-7, the system comprising: A distance matrix establishment module is used to determine the target warehouse, which has multiple cargo storage areas and multiple platforms, and to establish a distance matrix between the multiple cargo storage areas and multiple platforms. The arrival time prediction module is used to obtain the vehicle location information and operation demand information of multiple transport vehicles in real time, and predict the arrival time of the multiple vehicles based on the vehicle location information of the multiple transport vehicles. The matching conflict detection module is used to generate an initial vehicle platform matching scheme based on the distance matrix and the operation requirement information, and to perform conflict detection on the initial vehicle platform matching scheme based on the arrival times of the multiple vehicles, and obtain the conflict detection results. The matching scheme optimization module is used to optimize the initial vehicle platform matching scheme according to the distance matrix when there are conflicting vehicle platform pairs in the conflict detection results, and generate an optimized vehicle platform matching scheme for platform management in the logistics park.

Images