Beidou-based manufacturing industry contract logistics data fusion and scheduling method and system

Abstract

The invention discloses a Beidou-based manufacturing industry contract logistics data fusion and scheduling method and system, and belongs to the field of data processing methods specially suitable for management. In the method, a basic geography and road network data module obtains basic geography data; the Beidou dynamic monitoring data module collects transport vehicle data; the cargo and contract data module stores cargo and contract data; the logistics node data module obtains logistics node data; the data fusion processing module performs spatial matching on the basic geographic data, the transport vehicle data, the cargo and contract data and the logistics node data to obtain a manufacturing industry logistics data association table; the data fusion processing module performs data fusion based on the manufacturing industry logistics data association table to obtain a unified data set; and the intelligent scheduling module generates an optimal scheduling control instruction based on the unified data set and a preset constraint condition, and sends the optimal scheduling control instruction to the corresponding vehicle-mounted terminal. The method and the device are used for improving the accuracy of a scheduling scheme in a complex manufacturing industry scene.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of data processing methods specially applicable to management, and particularly relates to a manufacturing contract logistics data fusion and scheduling method and system based on Beidou. BACKGROUND

[0002] Manufacturing contract logistics refers to the logistics activity of transporting raw materials, semi-finished products or finished products of manufacturing enterprises to designated locations according to contract agreements. Traditional manufacturing contract logistics mainly relies on manual scheduling, and it is difficult to real-time grasp the vehicle position, cargo status and logistics node operation, resulting in low logistics efficiency and high cost.

[0003] The related technology proposes a manufacturing logistics scheduling system based on GPS. The system includes a GPS positioning module, an electronic map module and a scheduling module. The GPS positioning module is used to obtain vehicle position information; the electronic map module is used to store road network data; the scheduling module performs path planning and scheduling optimization according to the vehicle position information and the road network data. The system can improve the logistics transportation efficiency by real-time acquisition of vehicle position information and path optimization.

[0004] However, the core capability of the above-mentioned system is limited to tracking the physical state of the transportation tool (vehicle), and the data dimension processed by the system is essentially single, only containing geographic position and road network information. When the dispatcher sees the real-time position of the vehicle in the system, it is difficult to know the associated contract requirements, cargo status and delivery time window and other key business information behind it, because these information are physically isolated in independent business systems such as enterprise resource planning (ERP) and warehouse management system (WMS). The separation of business information and transportation execution information makes the decision information incomplete due to the single data dimension of the system, thereby reducing the accuracy of the scheduling scheme in complex manufacturing scenarios. SUMMARY

[0005] The application provides a manufacturing contract logistics data fusion and scheduling method and system based on Beidou, which is used to improve the accuracy of the scheduling scheme in complex manufacturing scenarios.

[0006] In a first aspect, the application provides a manufacturing contract logistics data fusion and scheduling method based on Beidou. The basic geographic and road network data module obtains basic geographic data; the Beidou dynamic monitoring data module collects transportation vehicle data; the cargo and contract data module stores cargo and contract data; the logistics node data module obtains logistics node data; The data fusion processing module takes Beidou positioning coordinates as the space-time reference to perform spatial matching on the basic geographic data, transportation vehicle data, cargo and contract data and logistics node data, and obtains a manufacturing logistics data association table; The data fusion processing module fuses the basic geographic data, the transportation vehicle data, the cargo and contract data and the logistics node data based on a manufacturing logistics data correlation table to obtain a unified data set; The intelligent scheduling module generates an optimal scheduling control instruction based on the unified data set and a preset constraint condition, and sends the optimal scheduling control instruction to a corresponding vehicle terminal.

[0007] By adopting the above technical solutions, various types of data are respectively acquired by the basic geographic and road network data module, the Beidou dynamic monitoring data module, the cargo and contract data module and the logistics node data module, and the data fusion processing module performs spatial matching and data fusion based on Beidou positioning coordinates as a space-time reference, so that logistics-related data scattered in different systems can be integrated into a unified data set. The unified data set contains complete logistics transportation chain information, and the scheduling control instruction generated by the intelligent scheduling module based on these information has higher accuracy and practicality. Based on the high-precision positioning capability of the Beidou system, the system can real-time master the position of the transportation vehicle, plan the path in combination with the road network data, and dynamically adjust the transportation scheme according to the cargo characteristics, contract requirements and logistics node state. This data fusion-driven scheduling mode can improve the accuracy of the scheduling scheme in complex manufacturing scenarios.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the basic geographic and road network data module includes a geographic and road network information module, a format conversion module, a coordinate conversion module, a fusion edge processing module and a data fusion module, and the step of acquiring the basic geographic data by the basic geographic and road network data module specifically includes: The geographic and road network information module acquires basic geographic information and road network information data; The format conversion module performs format uniform processing on the basic geographic information and road network information data to obtain uniform format data; The coordinate conversion module takes a Beidou high-precision control network as a spatial data basis, and converts the uniform format data into a preset geodetic coordinate system to obtain coordinate uniform data; The fusion edge processing module performs consistency processing on the geometric information and attribute information of the ground objects in the overlapping area in the coordinate uniform data to obtain consistency data; The data fusion module extracts transportation node data in the consistency data and performs fusion processing to obtain the basic geographic data.

[0009] By adopting the technical scheme, the basic geographic and road network data module processes the geographic information and road network information of different sources in a unified format through the format conversion module, solving the data heterogeneity problem. The coordinate conversion module uses the Beidou high-precision control network as the spatial data basis, and calculates the unified format data into the preset geodetic coordinate system, improving the accuracy and consistency of the spatial data. The fusion edge processing module processes the geometric information and attribute information of the ground objects in the overlapping area consistently, reducing data redundancy. The data fusion module extracts and fuses the transportation node data, constructs a complete transportation network, and improves the accuracy of the basic geographic data.

[0010] In combination with some embodiments of the first aspect, in some embodiments, the Beidou dynamic monitoring data module includes a Beidou dynamic monitoring information module, a first data preprocessing module, a first data clustering fusion module, a first data correlation module, and a first data display module. The step of collecting transportation vehicle data by the Beidou dynamic monitoring data module specifically includes: The Beidou dynamic monitoring information module collects original vehicle data of the transportation vehicle through the vehicle-mounted Beidou terminal; The first data preprocessing module performs noise reduction processing on the original vehicle data through nonlinear filtering to obtain preprocessed data; The first data clustering fusion module performs spatio-temporal clustering on the preprocessed data to obtain vehicle running track data; The first data correlation module inputs the vehicle running track data into a preset frequent correlation model to obtain transportation vehicle data, and the first data display module displays the transportation vehicle data.

[0011] By adopting the technical scheme, the Beidou dynamic monitoring data module uses nonlinear filtering to perform noise reduction processing on the original vehicle data, removing positioning noise and abnormal data. By processing the preprocessed data through the spatio-temporal clustering method, the system can identify and restore the actual running track of the vehicle, avoiding the deviation of the track caused by positioning errors. Inputting the vehicle running track data into the preset frequent correlation model can find the regularity features and correlation in the vehicle running process. This data processing method improves the accuracy and reliability of the vehicle position data.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the goods and contract data module includes a goods and contract information module, a second data preprocessing module, a data storage module, and a second data display module. The step of storing goods and contract data by the goods and contract data module specifically includes: The goods and contract information module collects goods electronic waybill information and contract electronic documents to obtain original business data; The goods and contract information module extracts goods information and contract information from the original business data; The second data preprocessing module performs standardized processing on the cargo information and the contract information to obtain cargo and contract data; The data storage module stores the cargo and contract data, and the second data display module displays the cargo and contract data.

[0013] By using the above technical solutions, the cargo and contract data module performs standardized processing on the information extracted from the cargo electronic waybill information and the contract electronic document to establish a standardized cargo and contract database. The standardized processing eliminates the format differences and data irregularities in the original business data. The data storage module centrally stores and manages the standardized data, facilitating quick retrieval and calling of related information by the system. This information processing method improves the standardization and usability of business data, enabling the system to accurately grasp the cargo transportation requirements and contract requirements and ensure that the scheduling scheme meets the actual business requirements.

[0014] In combination with some embodiments of the first aspect, in some embodiments, the logistics node data module includes a logistics node information module, a third data preprocessing module, a third data clustering fusion module, a third data correlation module, and a third data display module. The logistics node data module obtains logistics node data by specifically including the following steps: The logistics node information module collects node static data and operation state information of the logistics node; The third data preprocessing module performs standardized processing on the node static data to obtain cleaned data. The standardized processing includes removing duplicate data, supplementing missing data, and unifying data formats; The third data clustering fusion module classifies and aggregates the cleaned data based on geographic location and operation capacity level to obtain clustered data; The third data correlation module correlates the operation state information and the clustered data to obtain logistics node data, and the third data display module displays the logistics node data.

[0015] By adopting the technical scheme, the logistics node data module collects node static data and operation state information through the logistics node information module to obtain basic attributes and real-time operation conditions of the logistics node. The third data preprocessing module performs standardized processing on the node static data to remove duplicate data, supplement missing data and unify data formats, thereby improving data quality. The third data clustering fusion module classifies and aggregates the cleaned data based on geographical positions and operation capacity levels, so that the system can identify distribution characteristics of logistics nodes in different regions and with different capacity levels. The third data correlation module correlates the operation state information with the clustered data to form a complete data set containing static attributes and dynamic states of the logistics node. This data processing manner enables the system to accurately grasp geographical distribution, operation capacity and real-time state of each logistics node, which helps the system to fully consider actual operation conditions of the logistics node when making scheduling decisions, avoids scheduling goods to nodes with insufficient operation capacity or abnormal operation state, and improves precision of the scheduling scheme in complex manufacturing scenarios.

[0016] In combination with some embodiments of the first aspect, in some embodiments, the data fusion processing module includes a spatial matching module, a data attribute correlation module, a data configuration module, a data slice deployment module and a first service publishing module. The data fusion processing module takes Beidou positioning coordinates as a space-time reference to perform spatial matching on the basic geographic data, the transportation vehicle data, the goods and contract data and the logistics node data to obtain a manufacturing logistics data association table. The data fusion processing module performs data fusion on the basic geographic data, the transportation vehicle data, the goods and contract data and the logistics node data based on the manufacturing logistics data association table to obtain a unified data set. Specifically, the step includes: The spatial matching module establishes a space-time reference system taking Beidou positioning coordinates as a space-time reference; The spatial matching module maps the basic geographic data, the transportation vehicle data, the goods and contract data and the logistics node data to the space-time reference system to obtain a mapped space-time reference system; The data attribute correlation module performs attribute correlation on the basic geographic data, the transportation vehicle data, the goods and contract data and the logistics node data based on the mapped space-time reference system to obtain the manufacturing logistics data association table; The data configuration module merges data with an association relationship according to the manufacturing logistics data association table to obtain a fusion data set; The data slice deployment module performs data slicing on the fusion data set according to a preset business requirement, deploys the fusion data set after data slicing to a corresponding business server to obtain a unified data set, and publishes the unified data set by the first service publishing module.

[0017] By adopting the technical scheme, the data fusion processing module establishes a space-time reference system taking the Beidou positioning coordinates as a reference through the space matching module, and maps various types of data into a unified space-time framework. The data attribute association module associates the attributes of different types of data based on the space-time reference system, and establishes the logical relationship between the data. The data configuration module combines the data with associated relationships, eliminating the data island phenomenon. The data slice deployment module deploys the fused data set according to business needs, improving data access efficiency. This data fusion processing method solves the problems of scattered, heterogeneous, and poor correlation of logistics data, and realizes the deep fusion of basic geographic data, transportation vehicle data, cargo and contract data, and logistics node data in a unified space-time reference framework. Through data slice deployment and service publishing, the system can quickly obtain and process the required data information, improving data usage efficiency.

[0018] In combination with some embodiments of the first aspect, in some embodiments, the intelligent scheduling module includes a fusion information module, a scheduling matching module, a dynamic adjustment module, a scheme generation module, and a second service publishing module. The intelligent scheduling module generates an optimal scheduling control instruction based on the unified data set and the preset constraint condition, and sends the optimal scheduling control instruction to the corresponding vehicle terminal. Specifically, the step includes: The fusion information module reads the vehicle location information, cargo information, road network information, and node information in the unified data set, and performs data preprocessing and format conversion to obtain scheduling basic data; The scheduling matching module performs multi-objective optimization calculation on the scheduling basic data according to the preset constraint condition to obtain an initial scheduling scheme; The dynamic adjustment module optimizes and adjusts the initial scheduling scheme based on the obtained dynamic data to obtain an optimized scheduling scheme; The scheme generation module converts the optimized scheduling scheme into a standard scheduling instruction format to generate an optimal scheduling control instruction; The second service publishing module distributes the optimal scheduling control instruction to the corresponding vehicle terminal.

[0019] By adopting the technical solutions, the intelligent scheduling module pre-processes and converts the formats of various types of information in the unified data set through the fusion information module, improving the usability of the data. The scheduling matching module performs multi-objective optimization calculation based on preset constraint conditions, and can find an optimal scheduling scheme under the premise of meeting various types of constraints. The dynamic adjustment module optimizes and adjusts the initial scheduling scheme according to the real-time acquired dynamic data, so that the scheduling scheme can adapt to the dynamic changes of the logistics environment. The scheme generation module converts the optimized scheduling scheme into a standard scheduling instruction format, ensuring the standardization and executability of the scheduling instruction. This intelligent scheduling method can comprehensively consider various factors such as vehicle position, cargo characteristics, road network conditions, and node state, generate a more scientific and reasonable scheduling scheme, and can dynamically adjust according to real-time conditions, improving the accuracy of the scheduling scheme in complex manufacturing scenarios.

[0020] In a second aspect, the embodiments of the present application provide a Beidou-based manufacturing contract logistics data fusion and scheduling system, which comprises one or more processors and a memory; the memory is coupled with the one or more processors, and is used to store computer program codes, the computer program codes comprising computer instructions, and the one or more processors invoke the computer instructions to enable the system to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0021] In a third aspect, the embodiments of the present application provide a computer readable storage medium comprising instructions, which, when executed on a system, cause the system to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0022] In a fourth aspect, the embodiments of the present application provide a computer program product, which, when executed on a system, causes the system to perform the method described in any possible implementation manner of the first aspect.

[0023] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. The application provides a manufacturing contract logistics data fusion and scheduling method based on Beidou. Through the basic geographic and road network data module, the Beidou dynamic monitoring data module, the cargo and contract data module, and the logistics node data module, various types of data are obtained, and the data fusion processing module is used to perform spatial matching and data fusion based on Beidou positioning coordinates as the space-time reference, which can integrate logistics-related data scattered in different systems into a unified data set. The unified data set contains complete logistics transportation chain information, and the intelligent scheduling module generates scheduling control instructions based on these information with higher accuracy and practicality. Based on the high-precision positioning capability of the Beidou system, the system can real-time master the position of the transportation vehicle, plan the path combined with the road network data, and dynamically adjust the transportation scheme according to the characteristics of the goods, contract requirements, and logistics node state. This data fusion-driven scheduling method can improve the accuracy of the scheduling scheme in complex manufacturing scenarios.

[0024] 2. The application provides a manufacturing contract logistics data fusion and scheduling method based on Beidou. The data fusion processing module establishes a space-time reference system based on Beidou positioning coordinates through the spatial matching module, and maps various types of data into a unified space-time framework. The data attribute association module associates the attributes of different types of data based on the space-time reference system, establishing a logical relationship between the data. The data configuration module combines data with associated relationships, eliminating the phenomenon of data islands. The data slicing deployment module deploys the fused data set according to business needs, improving data access efficiency. This data fusion processing method solves the problem of scattered, heterogeneous, and poor correlation of logistics data, and realizes the deep fusion of basic geographic data, transportation vehicle data, cargo and contract data, and logistics node data in a unified space-time reference framework. Through data slicing deployment and service publishing, the system can quickly obtain and process the required data information, improving data usage efficiency.

[0025] 3. The application provides a manufacturing contract logistics data fusion and scheduling method based on Beidou. The intelligent scheduling module pre-processes and formats the various types of information in the unified data set through the fusion information module, improving the usability of the data. The scheduling matching module performs multi-objective optimization calculation based on preset constraints, which can find a better scheduling scheme under the premise of meeting various constraints. The dynamic adjustment module optimizes and adjusts the initial scheduling scheme based on real-time dynamic data, so that the scheduling scheme can adapt to the dynamic changes of the logistics environment. The scheme generation module converts the optimized scheduling scheme into a standard scheduling instruction format, ensuring the standardization and executability of the scheduling instruction. This intelligent scheduling method can consider multiple factors such as vehicle location, cargo characteristics, road network status, and node state, generate a more scientific and reasonable scheduling scheme, and dynamically adjust according to real-time conditions, improving the accuracy of the scheduling scheme in complex manufacturing scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a flowchart of one embodiment of a manufacturing contract logistics data fusion and scheduling method based on Beidou.

[0027] Figure 2 is another flowchart of one embodiment of a manufacturing contract logistics data fusion and scheduling method based on Beidou.

[0028] Figure 3 is a module structure diagram of a manufacturing contract logistics data fusion and scheduling system based on Beidou provided by an embodiment of the present application.

[0029] Figure 4 is a module structure diagram of a basic geographic and road network data module provided by an embodiment of the present application.

[0030] Figure 5 is a module structure diagram of a Beidou dynamic monitoring data module provided by an embodiment of the present application.

[0031] Figure 6 is a module structure diagram of a cargo and contract data module provided by an embodiment of the present application.

[0032] Figure 7 is a module structure diagram of a logistics node module provided by an embodiment of the present application.

[0033] Figure 8 is a module structure diagram of a data fusion processing module provided by an embodiment of the present application.

[0034] Figure 9 is a module structure diagram of an intelligent scheduling module provided by an embodiment of the present application.

[0035] Figure 10 is an entity device structure diagram of a manufacturing contract logistics data fusion and scheduling system based on Beidou provided by an embodiment of the present application. DETAILED DESCRIPTION

[0036] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to be limiting of the present application. As used in the specification and the appended claims of the present application, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term "and / or" as used herein refers to any or all possible combinations of one or more of the listed items.

[0037] Hereinafter, the terms "first", "second", "third", "fourth", "fifth", "sixth", "seventh" and "eighth" are used only for descriptive purposes and should not be construed as implying or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second", "third", "fourth", "fifth", "sixth", "seventh" and "eighth" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0038] In the context of today's deep integration of globalization and informatization, manufacturing is undergoing a profound transformation from traditional production mode to intelligent manufacturing and service-oriented manufacturing. The efficiency, resilience and intelligence level of the supply chain have become key indicators of measuring the core competitiveness of enterprises. Among them, manufacturing contract logistics, as the artery connecting the core links of raw material supply, production and processing, and finished product distribution, its operation efficiency directly affects the cost and response speed of the entire supply chain. Enterprises no longer pursue only point-to-point goods displacement, but need a smart logistics system that can deeply perceive, accurately predict and intelligently decide, in order to cope with the rapid changes in market demand, dynamic adjustments in production plans and increasingly complex delivery requirements.

[0039] Under this background, the related technology proposes a logistics scheduling system based on GPS. This system uses GPS technology to obtain the geographic position of the transport vehicle in real time, combines with the road network information in the electronic map, and plans the path through the scheduling algorithm. This technology to some extent gets rid of the traditional scheduling mode which relies on manual telephone communication and experience judgment, realizes the preliminary digital tracking of the transport tool, and improves the in-transit transparency of the vehicle.

[0040] However, when this technology is placed in a typical modern manufacturing application scenario, its inherent limitations become apparent. Imagine a large precision electronics manufacturer whose production process involves sourcing hundreds of high-value components globally, with products needing timely delivery to multiple assembly plants or customers located in different regions. This manufacturer's logistics activities are characterized by extremely high timeliness requirements (e.g., Just-In-Time (JIT) supply), high-value and diverse goods (requiring special transportation conditions), and complex contract terms (involving multiple carriers, delivery windows, penalties for breach of contract, etc.). In this scenario, a dispatcher using an existing GPS system can only see a single dot moving across the map representing a vehicle. It is difficult to intuitively determine whether the vehicle is carrying urgent materials that will cause production line shutdowns, or whether the warehouse the vehicle is expected to arrive at is currently full or the loading / unloading area is busy, preventing immediate operation. Key business information such as the latest delivery time stipulated in the contract and the required temperature and humidity environment for the goods is completely disconnected from the vehicle's physical operating status. This information silo effect caused by the single data dimension makes scheduling decisions severely lack contextual awareness, making it difficult to make truly optimal judgments under complex constraints, thereby reducing the accuracy of scheduling schemes in complex manufacturing scenarios.

[0041] To address the technical problem of insufficient accuracy in scheduling decisions caused by the separation of business information and physical execution information, this application provides a BeiDou-based method for data fusion and scheduling of manufacturing contract logistics. This method is applied to a system comprising a basic geographic and road network data module, a BeiDou dynamic monitoring data module, a cargo and contract data module, a logistics node data module, a data fusion processing module, and an intelligent scheduling module. Please refer to the schematic diagram of the module structure of the BeiDou-based manufacturing contract logistics data fusion and scheduling system provided in this application. Figure 3 This method constructs a data foundation centered on the BeiDou high-precision spatiotemporal reference, deeply integrating multi-source heterogeneous data such as geographic road networks, vehicle dynamics, cargo contracts, and logistics nodes to form a unified dataset with complete information and rich dimensions. Based on this dataset, intelligent algorithms are used to generate and dynamically optimize scheduling instructions. The following will elaborate on this method with examples.

[0042] The following example is used in conjunction with Figure 1 This application describes a BeiDou-based method for data fusion and scheduling of manufacturing contract logistics. Please see Figure 1 This is a flowchart illustrating a BeiDou-based manufacturing contract logistics data fusion and scheduling method in an embodiment of this application.

[0043] S101, the basic geographic and road network data module acquires basic geographic data; the Beidou dynamic monitoring data module collects transportation vehicle data; the cargo and contract data module stores cargo and contract data; the logistics node data module acquires logistics node data; The basic geographic and road network data module obtains basic geographic data, wherein the basic geographic and road network data module comprises a geographic and road network information module, a format conversion module, a coordinate conversion module, a fusion edge processing module and a data fusion module. Specifically, the geographic and road network information module obtains basic geographic information and road network information data; the format conversion module performs format uniform processing on the basic geographic information and road network information data to obtain uniform format data; the coordinate conversion module takes the Beidou high-precision control network as spatial data basis, and calculates the uniform format data into a preset geodetic coordinate system to obtain coordinate uniform data; the fusion edge processing module performs consistency processing on the geometric information and attribute information of the ground objects in the overlapping area of the coordinate uniform data to obtain consistency data; and the data fusion module extracts the transportation node data in the consistency data and performs fusion processing to obtain the basic geographic data. First, the geographic and road network information module serves as a data entrance and is responsible for obtaining original geographic spatial data and road network data from various channels (such as commercial map suppliers, open source map data such as OpenStreetMap, or national surveying and mapping institutions). Due to the different data sources, the data formats, coordinate systems and qualities of the data are different. Therefore, the format conversion module intervenes to convert all these heterogeneous data (such as Shapefile, GeoJSON, KML and the like) into a standard format uniform in the system for subsequent processing. The next key precision uniformization link is the coordinate conversion module, which takes the Beidou high-precision control network as a reference to convert the coordinates of all data into a “preset geodetic coordinate system” (usually referring to the CGCS2000 national geodetic coordinate system). This step ensures that the positions of all spatial elements have a uniform high-precision reference, eliminating the positioning deviation caused by the mixed use of different coordinate systems. After processing the internal problems of a single data, the fusion edge processing module solves the external problems in the splicing of multiple data. When data from different maps or sources overlap at the boundary, the module checks and corrects the geometric discontinuity or misplacement of “ground objects” such as roads, and the inconsistency in “attributes” (such as road names, grades), ensuring seamless splicing of the map and continuity of the information. Finally, the data fusion module extracts the transportation node data (such as intersections, toll stations, service areas, bridges and tunnels) that are crucial to logistics business from the clean map, and may enhance the attributes or build the topological relationship, and finally generates basic geographic data of high quality for the entire system. In this process, the specific types of geographic information data obtained and the specific technical means of data fusion processing can be adjusted according to the precision requirements and cost budget of the actual application scene, which is not limited here. Preferably, for the coordinate conversion module, a specific implementation manner is to utilize the open source geographic spatial data abstraction library (GDAL / OGR) and the built-in PROJ coordinate conversion engine.In the program, the meta information of the original data can be read to determine its source coordinate reference system (Source CRS), and then the target coordinate reference system (Target CRS) is set to the Beidou-based CGCS2000 (its EPSG code is 4490). By calling the coordinate conversion function provided by the library, the coordinates of each point in the data are accurately mathematically transformed, thereby completing the reduction. Another implementation manner is that for the system already deployed in the cloud, the geospatial service API provided by the cloud service provider can be called. The data containing the original coordinates are batch uploaded or streamed to the service, and the service returns the result after completing the coordinate conversion in the backend. This way outsources the complex coordinate system parameter maintenance and calculation tasks, simplifying the development and maintenance of the system itself. For the consistency processing of the fusion edge processing module, preferably, an automatic processing engine based on topological rules can be used. For example, the rules "at the map boundary, if the endpoints of two linear features (roads) are less than 5 meters apart, then automatically merge by adsorption", and "if the string similarity (such as using Jaro-Winkler distance calculation) of the 'road name' field of two adjacent roads is greater than 0.9, then unify to one of the names". Through the pre-stored geometric and attribute rules, most of the edge problems can be automatically processed. The specific module diagram of the basic geographic and road network data module is described in the following. Figure 4 .

[0044] The Beidou dynamic monitoring data module collects transportation vehicle data, wherein the Beidou dynamic monitoring data module comprises a Beidou dynamic monitoring information module, a first data preprocessing module, a first data clustering fusion module, a first data correlation module and a first data display module. Specifically, the Beidou dynamic monitoring information module collects original vehicle data of the transportation vehicle through a vehicle-mounted Beidou terminal; the first data preprocessing module performs noise reduction processing on the original vehicle data through nonlinear filtering to obtain preprocessed data; the first data clustering fusion module performs spatio-temporal clustering on the preprocessed data to obtain vehicle running track data; the first data correlation module inputs the vehicle running track data into a preset frequent correlation model to obtain transportation vehicle data, and the first data display module displays the transportation vehicle data. The Beidou dynamic monitoring information module serves as an interface of hardware and network, and continuously collects original vehicle data from the Beidou terminal installed on each vehicle through wireless communication (such as 4G / 5G). These data not only include the core latitude, longitude, elevation and timestamp, but also often have speed, heading, satellite signal quality (such as PDOP value) and other information. However, due to factors such as urban canyon effect, tunnels and bad weather, the original data often contains noise and drift, and direct use will lead to track jitter and position distortion. Therefore, the first data preprocessing module adopts nonlinear filtering technology for processing. This is a signal processing method that establishes a mathematical model of vehicle motion and combines actual observation values to make optimal estimates of the true position and speed of the vehicle, thereby filtering out random errors and outputting smooth and more accurate preprocessed data. Then, the first data clustering fusion module performs deep processing on these clean point data. It performs spatio-temporal clustering, that is, it finds the clustering area of data points in both time and space dimensions. For example, when a series of consecutive points stay in an area within a range of tens of meters within a few minutes, the algorithm identifies it as a "stay event", and the track of points moving quickly between points is identified as a "travel section". In this way, discrete points are aggregated into vehicle running track data with business meaning. Finally, the first data correlation module uses a preset frequent correlation model to perform higher-level analysis on the track data. The model can be a pattern recognition algorithm for discovering the frequently-visited places of the vehicle (even if not marked in the address library), the usual routes or abnormal behavior patterns (such as deviation from the conventional route). It correlates isolated track data with vehicle identity, historical behavior, etc., generates transportation vehicle data, and visualizes the data on an electronic map by the first data display module. In this process, the strength of nonlinear filtering, the radius and time threshold of spatio-temporal clustering, and the specific rules of the correlation model can be configured according to business needs, which are not limited here. Preferably, for the first data preprocessing module to perform noise reduction through nonlinear filtering, a specific implementation is to use Unscented Kalman Filter (UKF).Compared with the extended Kalman filter (EKF), the UKF approximates the nonlinear function by an unscented transformation, without the need to compute the Jacobian matrix, and has higher accuracy and better stability for strong nonlinear systems such as vehicle motion. The algorithm captures the mean and covariance of the state distribution through a set of carefully selected "Sigma points" and propagates them through the nonlinear model to obtain more accurate state estimates. Another implementation is to use particle filtering (Particle Filter). This method uses a large number of random samples (particles) with weights to represent the posterior probability distribution of the vehicle state, and is particularly suitable for handling non-Gaussian noise and multi-modal distribution problems, such as when a vehicle exits a tunnel and signal recovery may have multiple possible locations, particle filtering can better express this uncertainty. For the first data clustering fusion module, spatiotemporal clustering is performed, and the document describes its functions. Preferably, the TRA-DBSCAN algorithm can be used, which is a variant of the DBSCAN algorithm for trajectory data. It first uses DBSCAN to identify the stopping points in the trajectory (areas with high spatiotemporal density), and then defines the sequence of trajectory points between two stopping points as a trajectory segment. In this way, a complete original trajectory can be clearly segmented into a semantic sequence of "stop-travel-stop", obtaining structured vehicle running trajectory data. The construction method of the preset frequent association model can be: one is based on the combination of classical data mining algorithms. First, use density clustering algorithms such as DBSCAN or OPTICS to cluster all "stopping points" accumulated over a long period of time, thereby automatically discovering "frequently visited locations". Then, the vehicle's travel trajectory is converted into a road segment ID sequence in the road network through a map matching algorithm (such as the Hidden Markov Model HMM-based algorithm). Finally, apply sequence pattern mining algorithms (such as GSP - Generalized Sequential Patterns or PrefixSpan) to mine frequent subsequences in these road segment ID sequences, which represent "customary routes". Abnormal behavior is identified by comparing real-time trajectories with the mined frequent locations and route patterns. Another more advanced implementation is to use a deep learning-based representation learning model, especially an LSTM (Long Short-Term Memory Network) autoencoder. The specific implementation process is as follows: First, use a large amount of normal historical trajectory data (which can be coordinate sequences, speed sequences, etc.) to train an LSTM autoencoder. The model learns how to compress a normal trajectory sequence into a low-dimensional vector (encoding), and can almost losslessly reconstruct the original trajectory from this vector. After training, the model has mastered the internal patterns of "normal behavior" of the vehicle.In real-time monitoring, new trajectory data is input into this trained model. If the model can reconstruct the trajectory with a small "reconstruction error", it means that it is consistent with the historical normal pattern. On the contrary, if the reconstruction error is large, it means that the trajectory is "unseen" by the model, which can be determined as "abnormal behavior pattern". By analyzing the structure of this low-dimensional vector space, trajectories can also be clustered to discover patterns such as habitual routes. Please refer to the module schematic diagram of the Beidou dynamic monitoring data module. Figure 5 .

[0045] The cargo and contract data module stores cargo and contract data, wherein the cargo and contract data module comprises a cargo and contract information module, a second data preprocessing module, a data storage module, and a second data display module. Specifically, the cargo and contract information module collects cargo electronic waybill information and contract electronic documents to obtain original business data; the cargo and contract information module extracts cargo information and contract information from the original business data; the second data preprocessing module standardizes the cargo information and the contract information to obtain cargo and contract data; the data storage module stores the cargo and contract data, and the second data display module displays the cargo and contract data. The cargo and contract data module obtains "cargo electronic waybill information" and "contract electronic documents" from multiple source systems (such as ERP, CRM, WMS, TMS, or even email attachments or shared folders) in the enterprise through API interface, database direct connection or file analysis, etc. These constitute the original business data. These original data have various forms, including highly structured data such as electronic waybills and a large amount of unstructured or semi-structured text such as contract documents. Therefore, the module also needs to perform preliminary extraction work, using information extraction technology to identify and extract key fields from contract documents (such as PDF, Word files), such as contract number, parties, delivery address, delivery time window, default clause, and special requirements for goods (such as temperature and humidity control). Next, the second data preprocessing module standardizes the extracted cargo information and contract information. This includes data cleaning (such as removing duplicate waybill records), format unification (such as converting "2023-12-25" and "25 / 12 / 2023" to "2023-12-25"), unit conversion (such as converting pounds to kilograms), and code mapping (such as converting different customer names to a unique customer ID). Through this series of processing, the originally chaotic data is transformed into consistent format and clear semantic cargo and contract data. Subsequently, the data storage module persists these standardized data in a dedicated database for efficient querying and management. Finally, the second data display module provides a user interface for relevant personnel (such as sales, customer service, and dispatchers) to conveniently view and retrieve this information, for example, by waybill number to query the detailed attributes of the goods and the associated contract terms. In this process, the accuracy of information extraction and the specific rules of data standardization can be customized according to the characteristics of the enterprise business process, which is not limited here. Preferably, for the cargo and contract information module to extract information from contract electronic documents, one specific implementation is to use information extraction technology based on template matching and regular expressions. For contract documents with relatively fixed formats, a template can be predefined for each type of contract, with the fixed text pattern of the key information (such as "latest delivery date:") and its approximate position in the document.The system first identifies the type of the new contract and then applies the corresponding template to match and extract the values of the key fields through regular expressions. This method is simple to implement and works well for formatted documents. A more powerful and flexible implementation is to use natural language processing (NLP)-based named entity recognition (NER) technology. A set of annotated contract documents (i.e., manually labeled with entities such as "delivery address," "goods name," etc.) can be prepared in advance, and then a deep learning model (such as BERT-NER or CRF-LSTM) is trained using these data. After training, the model can understand the contextual semantics of the contract text and automatically identify and extract various predefined entities, even if they are expressed differently in different contracts. This method is more adaptable to unstructured text. For the second data preprocessing module, standardization is performed, and preferably, a master data management (MDM) system can be built. This system maintains a set of enterprise-level "golden standard" data, such as customer master data, product master data, and location master data. During preprocessing, all data extracted from business systems is compared and cleaned with the master data. For example, through fuzzy matching algorithms, "ABC Limited Company" and "ABC Company" are mapped to the unique customer ID in the master data, ensuring data consistency. The module diagram of the goods and contract data module is shown in the following figure. Figure 6 .

[0046] The logistics node data module acquires logistics node data, wherein the logistics node data module comprises a logistics node information module, a third data preprocessing module, a third data clustering fusion module, a third data correlation module, and a third data display module. Specifically, the logistics node information module collects node static data and operation state information of the logistics node; the third data preprocessing module performs standardization processing on the node static data to obtain cleaned data, and the standardization processing includes removing duplicate data, supplementing missing data, and unifying data formats; the third data clustering fusion module classifies and aggregates the cleaned data based on geographical position and operation capacity level to obtain clustered data; the third data correlation module correlates the operation state information and the clustered data to obtain logistics node data, and the third data display module displays the logistics node data. The logistics node information module is responsible for collecting two types of core data from multiple channels: one is node static data, which refers to attributes that are long-term stable and slow-changing, such as the geographical coordinates, address, warehouse area, shelf capacity, loading and unloading platform quantity, operation time window, and equipped device type and quantity of the node; the other is operation state information, which refers to dynamic data that changes in real time or quasi-real time, such as current inventory level, available storage location, number of vehicles waiting for loading and unloading, and platform occupancy. These data can come from WMS (warehouse management system), reservation queuing system, Internet of Things sensors (such as truck scales and access control), etc. Due to the diversity of data sources, the third data preprocessing module first performs standardization processing on the relatively stable node static data. This process includes: removing duplicate data by comparing node names and addresses; supplementing missing data (such as the operation time of a node not being filled in) using historical data or default values; and unifying data formats (such as unifying area units to square meters). After processing, a clean and complete cleaned data is obtained. Next, the third data clustering fusion module classifies and aggregates these cleaned nodes. It mainly depends on two dimensions: geographical position and operation capacity level. For example, the system can aggregate multiple small warehouses that are geographically adjacent (such as within the same industrial park) and have similar operation capacity (such as both having cold chain storage and loading and unloading capacity) into a regional capacity center at the macro planning level. This aggregation helps to simplify the network structure and facilitate regional resource coordination. Finally, the third data correlation module correlates dynamic operation state information with aggregated static clustered data in real time. This means that the system not only knows the basic capacity of a node (or node cluster), but also real-time monitors its current busy degree and available resources. For example, the dynamic information "A warehouse currently has 5 vehicles in queue" is correlated to the static data "A warehouse (has 10 loading and unloading platforms)". The finally generated logistics node data is a complete view containing static capacity and dynamic state, and is visualized on a map or dashboard by the third data display module, providing a direct basis for dispatching decisions.Preferably, the third data cluster fusion module classifies and aggregates based on geographical location and job capacity level. One specific implementation is to use a two-stage clustering method. In the first stage, use a geographical clustering algorithm (such as K-Means or DBSCAN) to cluster all node coordinates, and divide nodes that are geographically close into the same geographical cluster. In the second stage, within each geographical cluster, further cluster according to job capacity level. The job capacity level can be a multi-dimensional vector, such as [dry goods storage capacity, cold chain storage capacity, loading and unloading efficiency level,...]. Methods such as hierarchical clustering can be used to aggregate nodes with similar capabilities based on the similarity of these capability vectors. The final clustering result is a group of nodes that are both geographically concentrated and functionally similar. Another implementation is to define a weighted hybrid distance metric and use it in a single clustering algorithm. This distance metric function D(node_A, node_B) considers both geographical distance Dist_geo(A, B) and capability difference Diff_cap(A, B), for example D = w_geo x Dist_geo + w_cap x Diff_cap. Where w_geo and w_cap are weight coefficients, which can adjust the importance of geographical factors and capability factors according to business needs. Then, apply this custom distance metric to any clustering algorithm that supports custom distance (such as DBSCAN) to complete the aggregation based on location and capacity at one time. Please refer to the module diagram of the logistics node data module. Figure 7 .

[0047] S102, the data fusion processing module takes Beidou positioning coordinates as the space-time reference to perform spatial matching on the basic geographic data, transportation vehicle data, goods and contract data, and logistics node data, and obtains a manufacturing logistics data association table; S103, the data fusion processing module performs data fusion on the basic geographic data, transportation vehicle data, goods and contract data, and logistics node data based on the manufacturing logistics data association table, and obtains a unified data set; The step S102 and the step S103 are both executed by a data fusion processing module, wherein the data fusion processing module comprises a space matching module, a data attribute association module, a data configuration module, a data slice deployment module and a first service publishing module. Specifically, the space matching module establishes a space-time reference system with the Beidou positioning coordinates as the space-time reference; the space matching module maps the basic geographic data, the transport vehicle data, the goods and contract data and the logistics node data to the space-time reference system to obtain a mapped space-time reference system; the data attribute association module performs attribute association on the basic geographic data, the transport vehicle data, the goods and contract data and the logistics node data based on the mapped space-time reference system to obtain a manufacturing industry logistics data association table; the data configuration module merges the data with the association relationship according to the manufacturing industry logistics data association table to obtain a fusion data set; the data slice deployment module performs data slicing on the fusion data set according to a preset business requirement, deploys the fusion data set after the data slicing to a corresponding business server to obtain a unified data set, and publishes the unified data set by the first service publishing module.

[0048] Step S102 is the key bridge to realize data fusion, and its core task is "spatial matching", that is, to establish a correlation between the data from different modules, which are originally isolated, with the time and space reference defined by the Beidou positioning coordinates. This step is executed by the spatial matching module and the data attribute association module in the data fusion processing module. First, establishing a time and space reference system with Beidou positioning coordinates as the time and space reference means that the system creates a unified and high-precision four-dimensional (three-dimensional space + one-dimensional time) coordinate frame, and the "origin" and "scale" of this frame are defined by the Beidou system, which ensures high geospatial accuracy and time synchronization. Subsequently, the spatial matching module "maps" all four types of data obtained in step S101 - basic geographic data (such as roads, points of interest), transportation vehicle data (such as real-time location, historical trajectory), goods and contract data (such as origin, destination, delivery address), and logistics node data (such as warehouse, factory location) - into this unified time and space reference system. The mapping process is essentially converting the geographic location information (whether it is latitude and longitude coordinates, address description, or node ID) contained in each type of data into standard coordinates in the reference system. For example, a text address "123, some street, some district, some city" will be converted into accurate Beidou coordinates through a geocoding service. After mapping is complete, the data attribute association module begins to work, which associates attributes based on the location relationship and proximity of data in the time and space reference system, and finally generates a "manufacturing logistics data association table". This association table is like an index directory, which records that at time T, the truck is located at coordinates (X, Y, Z), it is driving on 'G2 Expressway', the contract number it is carrying is 'C2023001', the contract corresponds to 'precision instruments', and the destination is the warehouse with ID 'WH005'. The way this association is established can be diverse, for example, the association between vehicle and road is based on the spatial inclusion or proximity relationship between vehicle coordinates and road geometry; the association between vehicle and goods is based on the shipping information; the association between goods and contract is based on the contract number. The granularity of the association can also be flexible, which is not limited here.

[0049] For the process of mapping the basic geographic data, transportation vehicle data, cargo and contract data, and logistics node data to the spatio-temporal reference system by the spatial matching module, a specific implementation is to build a spatial database (such as PostGIS or Oracle Spatial). The database natively supports geospatial data types and spatial indexes (such as R-Tree), and can efficiently store and query data with geographic coordinates. The mapping process is to store the cleaned and converted data in the corresponding table of the database. For example, the road data is stored as LineString type, the vehicle position is stored as Point type, and the logistics node is stored as Point type. For the address information in the cargo and contract data, it can be converted into coordinates by calling a geographic coding API (such as Gaode, Baidu Map API, or a self-built address library), and then stored in the database. Another implementation is to use an in-memory computing geospatial processing engine (such as GeoMesa or Apache Sedona). This way loads the geographic data into a distributed in-memory computing framework (such as Spark), uses its powerful parallel processing capability, and performs real-time spatial mapping and correlation calculation on massive stream data (such as vehicle trajectory) and static data (such as road network), with higher throughput and lower latency. For the data attribute association module, the basic geographic data, transportation vehicle data, cargo and contract data, and logistics node data are associated based on the mapped spatio-temporal reference system. The preferred way is to perform a series of spatial join queries. For example, by using spatial relationship functions such as ST_Intersects or ST_DWithin, the roads intersected or adjacent to the current vehicle position point are queried, so as to associate the road attributes (such as name, speed limit) with the vehicle. At the same time, through non-spatial connection, the vehicle and cargo information are associated according to the waybill number, and the cargo and contract terms are associated according to the contract number, and finally all the association relationships are materialized into the manufacturing logistics data association table.

[0050] Step S103 is the execution and landing phase of data fusion, the core task of which is to combine the physically dispersed data into a logically unified data set serving the business based on the association relationship established in step S102, and deploy it for calling by upper-layer applications. This step is completed by the data configuration, data slicing deployment and first service publishing modules in the data fusion processing module. First, the data configuration module performs data merging operation according to the manufacturing logistics data association table generated in the last step. This process can be understood as that the system splices the record rows with association relationship from different data tables according to the formula in the association table to form a wide table or a structured data object, which contains all relevant information of a logistics entity (such as a transport task), such as the real-time position and speed of the vehicle, the name, quantity, temperature and humidity requirements of the cargo, the delivery time window and urgency of the associated contract, the congestion status of the road and the operation capacity of the front logistics node, etc., thereby obtaining the fusion data set. Next, the data slicing deployment module processes the huge fusion data set. The so-called "data slicing" (or data slicing) is to extract a specific data subset from the fusion data set according to the pre-set business requirements. For example, the finance department may only need to concern the cost and mileage related financial slices, while the dispatch center needs to include the real-time position, ETA (estimated time of arrival) and node status operation slices. This slicing operation can optimize the efficiency of data transmission and processing, and ensure that each business unit only obtains the minimum necessary information required. Then, these data slices are deployed to the corresponding business servers, which means that the system pushes or stores the processed data to places that can be accessed by various business applications (such as dispatch system, monitoring large screen, financial system), forming the final unified data set. Finally, the first service publishing module exposes the unified data set in the form of standardized service interface (such as RESTful API, GraphQL or data stream) for calling by intelligent scheduling module or other authorized applications. The form of service can be diverse, such as providing request-response service according to vehicle ID to query its complete state, or providing subscription type data stream to "push all high-priority cargo vehicle position updates in real time", which is not limited here.

[0051] Preferably, for the process of merging data with association relationship according to the manufacturing logistics data association table by the data configuration module to obtain the fused data set, a specific implementation manner is to adopt an ETL (Extract-Transform-Load) tool or a data integration platform (such as Apache NiFi, Kettle) to construct a data pipeline. The pipeline is triggered by time or event, reads the manufacturing logistics data association table, and extracts corresponding data records from each source database (or data lake) according to the manufacturing logistics data association table, performs connection (Join) and aggregation operations in the memory, and finally writes the generated fused data set into a special data warehouse (such as ClickHouse, Apache Druid). The data warehouse is optimized for fast query and analysis of large-scale data. Another implementation manner is to adopt a virtual data integration technology (also known as data federation). In this manner, data is not physically moved and merged. Instead, a virtual intermediate layer is constructed, which contains the metadata and association rules of all source data. When the upper-layer application initiates a query, the virtual data integration engine decomposes the query in real time, sends it to each source data system, and then dynamically merges and processes the returned results to finally return them to the application. The advantage of this manner is that data always remains up-to-date and no additional storage space is required, but the query pressure on the source system is greater. For the data slice deployment module and the first service publishing module, preferably, a micro-service architecture can be constructed. Each data slice can be generated and maintained by a special micro-service. For example, a real-time location service is responsible for processing and publishing the slice related to the vehicle location, and a contract status service is responsible for publishing the slice related to the contract performance status. These micro-services provide services to the outside through an API gateway, realizing the decoupling and independent expansion of functions. The module schematic diagram of the data fusion processing module is shown in Figure 8 .

[0052] In S104, the intelligent scheduling module generates optimal scheduling control instructions based on the unified data set and preset constraint conditions, and sends the optimal scheduling control instructions to corresponding vehicle terminals.

[0053] The intelligent scheduling module generates optimal scheduling control instructions based on the unified data set and preset constraint conditions, and sends the optimal scheduling control instructions to corresponding vehicle terminals. The intelligent scheduling module includes a fusion information module, a scheduling matching module, a dynamic adjustment module, a scheme generation module, and a second service publishing module. Specifically, the fusion information module reads vehicle location information, cargo information, road network information, and node information in the unified data set, and performs data preprocessing and format conversion to obtain scheduling basic data. The scheduling matching module performs multi-objective optimization calculation on the scheduling basic data according to the preset constraint conditions to obtain an initial scheduling scheme. The dynamic adjustment module optimizes and adjusts the initial scheduling scheme based on the obtained dynamic data to obtain an optimized scheduling scheme. The scheme generation module converts the optimized scheduling scheme into a standard scheduling instruction format to generate optimal scheduling control instructions. The second service publishing module distributes the optimal scheduling control instructions to corresponding vehicle terminals.

[0054] First, the fusion information module serves as a data interface, responsible for reading various types of information required for scheduling from the unified data center, such as real-time vehicle positions, cargo attributes, real-time road conditions, and logistics node queuing situations, and performing necessary preprocessing and format conversion to form scheduling basis data that can be directly used by algorithms. Next, the scheduling matching module is the brain of intelligent scheduling. It performs multi-objective optimization calculations on scheduling basis data based on "preset constraints." Preset constraints include hard constraints (such as vehicle load limits, driver work time regulations, and delivery time windows that must be observed) and soft constraints (such as the desire for the lowest transportation cost, the shortest total distance, and the highest customer satisfaction). "Multi-objective optimization" means that the algorithm needs to find a best balance point among these possibly conflicting goals, for example, the fastest path may cost more. This calculation process produces an initial scheduling scheme, which is a comprehensive transportation plan that takes into account various factors and is currently the best. However, the logistics environment is dynamic, so the dynamic adjustment module continuously monitors dynamic data (such as new traffic congestion, weather changes, and customer temporary changes in delivery times). Once it finds that these changes may make the initial scheme no longer optimal or feasible, it triggers re-optimization or local adjustment to generate an optimized scheduling scheme. Finally, the scheme generation module converts this optimized scheme (which may be a series of path points and time points) into machine-readable standard scheduling instruction formats such as path planning coordinate sequences, recommended speeds, and next station instructions, generating optimal scheduling control instructions. The second service publishing module is responsible for accurately distributing these instructions to corresponding vehicle-mounted terminals through wireless communication networks (such as 4G / 5G), guiding drivers to execute them. Preferably, for the scheduling matching module to perform multi-objective optimization calculations on scheduling basis data based on preset constraints, the document describes its functions, and one specific implementation is to use heuristic optimization algorithms such as genetic algorithms (Genetic Algorithm) or ant colony algorithms (Ant Colony Optimization). Taking genetic algorithms as an example, a complete scheduling scheme (i.e., the allocation of paths for all vehicles) can be encoded as a "chromosome." The algorithm first randomly generates an initial "population" (i.e., a set of initial schemes), and then iterates through simulating selection, crossover, and mutation operations in biological evolution. In each generation, the fitness of each scheme is evaluated based on a fitness function that combines time, cost, punctuality, and other factors, and the winners are retained and reproduced for the next generation. After enough iterations, the individual with the highest fitness in the population is the approximate optimal solution found, which is the initial scheduling scheme. Another implementation is to use mixed integer programming (MIP) in precise optimization algorithms. The scheduling problem can be mathematically modeled as a set of linear equations and inequalities containing continuous variables (such as arrival times) and integer variables (such as whether a vehicle chooses a certain path).Then a commercial or open-source MIP solver (such as Gurobi, CPLEX, or SCIP) is used to solve it. This approach can find the theoretically optimal solution, but the computation time can be long. For the “dynamic adjustment module”, it is preferred to use an event-driven architecture. The system subscribes to change events of specific metrics in the unified dataset (such as “road condition becomes congested” or “ETA delay exceeds threshold”). Once an event is received, the dynamic adjustment module evaluates the impact of the event on the current dispatching scheme. For local impact (such as a single vehicle path being blocked), a fast path re-planning algorithm (such as a variant of A* algorithm) can be run for that vehicle only; for global impact (such as a key node being closed), a full re-dispatching can be triggered. The module diagram of the intelligent dispatching module is shown in Figure 9 .

[0055] In the above embodiments, various types of data are obtained by the basic geographic and road network data module, the Beidou dynamic monitoring data module, the cargo and contract data module, and the logistics node data module, respectively, and spatial matching and data fusion are performed by the data fusion processing module using Beidou positioning coordinates as the spatio-temporal reference, so that logistics-related data scattered in different systems can be integrated into a unified dataset. The unified dataset contains complete logistics transportation chain information, and the dispatching control instructions generated by the intelligent dispatching module based on these information have higher accuracy and practicality. Based on the high-precision positioning capability of the Beidou system, the system can real-time master the positions of the transportation vehicles, perform path planning in combination with road network data, and dynamically adjust the transportation scheme according to the cargo characteristics, contract requirements, and logistics node states. This data fusion-driven dispatching method can improve the accuracy of the dispatching scheme in complex manufacturing scenarios.

[0056] In the first embodiment, unified management of manufacturing contract logistics data is achieved by obtaining various types of data by the basic geographic and road network data module, the Beidou dynamic monitoring data module, the cargo and contract data module, and the logistics node data module, respectively, and performing spatial matching and data fusion by the data fusion processing module using Beidou positioning coordinates as the spatio-temporal reference. However, in the logistics transportation scenario in densely populated urban areas, when multiple transportation vehicles receive dispatching instructions under similar spatio-temporal conditions (such as multiple vehicles transporting in the same area at the same time during the morning rush hour), if these vehicles travel according to the respective dispatching instructions received, vehicle aggregation may occur in the local area, causing traffic microcirculation problems. To solve this technical problem, the intelligent dispatching module provided in the embodiments of the present application further includes an instruction timing simulation module and a collaborative optimization module, the timing simulation module includes a spatio-temporal scenario construction unit, an instruction execution simulation unit, and an aggregation situation analysis unit, and the collaborative optimization module includes an instruction interference evaluation unit, a timing rearrangement optimization unit, and a feedback learning adaptation unit. The following will be described in combination with Figure 2Another method for manufacturing contract logistics data fusion and scheduling based on Beidou in the embodiments of the application is described: Please refer to Figure 2 Another flowchart of a method for manufacturing contract logistics data fusion and scheduling based on Beidou in the embodiments of the application is shown.

[0057] S201, a space-time scene construction unit constructs a grid-based urban road space-time digital twin scene; The space-time scene construction unit constructs a grid-based urban road space-time digital twin scene, which means converting the real urban road network into a digital three-dimensional space-time model. The continuous geographic space is discretized into regular grid cells through the gridding method, and each grid cell carries specific space-time attribute information. This digital twin scene not only contains static road topological structure, lane information, and traffic facility distribution, but also integrates dynamic traffic flow, vehicle density, and signal light state, which are real-time changing space-time elements. The granularity of the grid can be adjusted according to the application precision requirements, usually set to 10m x 10m to 100m x 100m. Time slicing is involved in the modeling of space-time dimensions, which is generally discretized by minutes or seconds. The digital twin scene also needs to establish a coordinate system mapping relationship to accurately correspond the Beidou coordinate system and the grid coordinate system. Various thresholds in the scene model include traffic density threshold, vehicle speed threshold, congestion determination threshold, etc. The specific values of these thresholds can be personalized set according to the traffic characteristics and management requirements of different cities, which are not limited here.

[0058] Preferably, the grid-based space-time scene construction can adopt a multi-level modeling method. The bottom layer is the basic geographic information layer, including vector data of road network, elevation information, and land use type. The middle layer is the traffic facility layer, covering the spatial position and attribute information of static traffic elements such as signal lights, signs, and parking lots. The top layer is the dynamic traffic flow layer, which updates the vehicle position, speed, direction, and other motion state data in real time. In the specific implementation process, first, obtain the city basic geographic data and road network data, use GIS spatial analysis technology for gridding processing, then integrate Beidou positioning data to establish a dynamic updating mechanism, and finally construct a four-dimensional space-time model through time series modeling method. Another implementation method is to use deep learning-based scene reconstruction technology to automatically generate high-fidelity digital twin scenes by learning the space-time distribution pattern of historical traffic data through convolutional neural networks. This method can better capture the complex traffic flow dynamic characteristics and nonlinear space-time correlation.

[0059] In the process of constructing the space-time digital twin scene, the technical problem of data update delay may cause the scene model to be inconsistent with the actual situation. To solve this problem, a predictive update mechanism can be introduced, a state prediction model based on Kalman filter is established, the traffic state at the next time is predicted by using historical data and current observation data, and the scene model is updated in advance before the actual data arrives. The prediction model combines the vehicle kinematics model and the traffic flow theory, considers the constraint conditions such as road speed limit, traffic signal and road condition, calculates the possible position and state distribution of the vehicle at the future time, so as to maintain the real-time and accuracy of the digital twin scene.

[0060] S202, the instruction execution simulation unit projects the multiple scheduling instructions to be issued in time sequence into the digital twin scene; The instruction execution simulation unit projects the multiple scheduling instructions to be issued in time sequence into the digital twin scene refers to the process of pre-demonstrating and verifying the scheduling commands to be sent to each transport vehicle in the virtual environment. Each scheduling instruction contains key information such as vehicle identification, starting position, target position, estimated departure time, path planning, speed requirement, etc. Time sequence projection refers to simulating the execution of these scheduling tasks in the digital twin scene according to the time sequence of the instructions, observing the running track and mutual influence of each vehicle in the virtual environment. The projection process needs to consider the physical constraints of the vehicle, such as acceleration limit, turning radius, braking distance, etc., and the road constraints, such as lane limit, traffic rules, signal control, etc. During the simulation execution process, each scheduling instruction is converted into a series of space-time coordinate points to form the expected vehicle motion track. The time sequence arrangement of the instructions can be parallel, serial or mixed, and the specific arrangement method is determined according to factors such as the urgency of the logistics task, the availability of the vehicle, the road capacity, etc., which is not limited here.

[0061] Preferably, the time sequence projection can adopt a discrete simulation method based on event-driven, regarding each scheduling instruction as an event and processing them one by one in the simulation clock advancing process according to time priority. In specific implementation, an event queue is established to store all the scheduling instructions to be processed, the simulation engine takes out the instructions to be executed at the current time from the queue, generates the corresponding virtual vehicle in the digital twin scene, and calculates the motion track and state change of the vehicle according to the instruction parameters. The vehicle position, road occupancy, traffic flow state and other information are updated in real time during the simulation process, and the interaction and potential conflict between vehicles are detected. Another implementation method is to model each vehicle corresponding to a scheduling instruction as an intelligent agent, endow it with autonomous decision-making and environment perception ability, and perform distributed simulation in the digital twin scene. The intelligent agent makes path planning and behavior decision according to local environment information and global target, which can more realistically simulate the vehicle behavior in complex traffic scenes.

[0062] In the instruction timing projection process, the technical problem of simulation calculation complexity being too high to cause real-time deficiency can occur. To solve this problem, a hierarchical simulation architecture can be used to decompose a complex urban road network into multiple relatively independent sub-regions, and the scheduling instructions in each sub-region can be simulated in parallel, and only when a vehicle crosses the regional boundary, the state synchronization across regions is performed. This method reduces the calculation scale of a single simulation through spatial decomposition, and uses the parallel computing capability of a multi-core processor to improve the simulation efficiency, significantly shortens the calculation time under the premise of ensuring the simulation accuracy, and meets the timeliness requirements of real-time scheduling.

[0063] S203, an aggregation situation analysis unit identifies the vehicle aggregation situation appearing in the simulation process; The aggregation situation analysis unit identifying the vehicle aggregation situation appearing in the simulation process refers to timely discovering the vehicle aggregation phenomenon that can cause traffic congestion or transport efficiency reduction by analyzing the spatial distribution and density change of vehicles in the virtual simulation environment. The vehicle aggregation situation includes two dimensions of spatial aggregation and time aggregation. The spatial aggregation refers to that the vehicle density in a specific region exceeds the normal level, and the time aggregation refers to that the frequency of vehicles arriving at or passing through a region is abnormally high in a specific time period. The identification of the aggregation situation needs to establish a multi-level monitoring index system, including quantitative indexes such as regional vehicle density, vehicle speed distribution, vehicle spacing, and residence time. The judgment basis of the aggregation degree includes parameters such as density threshold, speed threshold, aggregation radius, and duration threshold. These thresholds can be dynamically adjusted according to different road types, time period characteristics, weather conditions, and other factors. The situation analysis also needs to consider the development trend of aggregation, predict the evolution direction of the aggregation phenomenon through time series analysis, and distinguish between temporary aggregation and persistent aggregation, as well as local aggregation and large-scale aggregation. The specific aggregation type classification standard and warning level setting are determined according to the management requirements of the actual application scene, which is not limited here.

[0064] Preferably, the vehicle aggregation situation identification can adopt a spatial analysis method based on density clustering, and use the DBSCAN algorithm to perform clustering analysis on the vehicle positions in the simulation scene to automatically identify high-density vehicle aggregation regions. In the specific implementation process, the position coordinates of all vehicles are sampled at fixed time intervals, the number of vehicles in each grid unit and the average speed are calculated, when the vehicle density exceeds the preset threshold and the average speed is lower than the normal level, the region is marked as a potential aggregation point, then adjacent aggregation points are merged into an aggregation region through connectivity analysis, and finally the geometric characteristics and influence range of the aggregation region are calculated. Another implementation manner is to use a network analysis method based on graph theory, model the road network as a graph structure, and take the vehicles as nodes on the graph, and identify the aggregation mode in the network by analyzing the node degree distribution and aggregation coefficient. This method can better consider the influence of road topological structure on vehicle aggregation, and is suitable for aggregation situation analysis in complex road network environment.

[0065] S204, the instruction interference evaluation unit constructs a space-time correlation graph between scheduling instructions; The instruction interference evaluation unit constructing a space-time correlation graph between scheduling instructions refers to establishing a network topology structure describing the mutual influence degree between instructions by analyzing the space-time interaction relationship of different scheduling instructions in the execution process. The space-time correlation graph takes scheduling instructions as nodes and interference relationships between instructions as edges, and the weight of the edge represents the interference strength or influence degree. The interference between instructions mainly reflects in space-time conflict, including path overlap, time conflict, resource competition, etc. Path overlap refers to the existence of common road sections of different instructions corresponding to vehicle driving routes, time conflict refers to multiple vehicles passing through the same area at similar times, and resource competition refers to multiple vehicles simultaneously needing to use limited road capacity or logistics nodes. The construction of the correlation graph needs to quantify the interference degree between instructions, and the commonly used measurement indexes include space-time distance, overlap degree, conflict probability, etc. The space-time distance comprehensively considers the spatial distance and time difference, the overlap degree reflects the sharing degree of the path or area, and the conflict probability predicts the possibility of conflict when the instruction is executed based on historical data and traffic model. The topology structure of the correlation graph can be an undirected graph or a directed graph, and the specific structure type and weight calculation method are selected according to the characteristics of the application scene and the analysis requirements, which are not limited here.

[0066] Preferably, the space-time correlation graph construction can adopt a method based on space-time trajectory analysis. First, according to the start and end points, the expected path and the time arrangement of each scheduling instruction, the corresponding space-time trajectory is generated, and then the space-time intersection and the minimum distance between trajectories are calculated. When the intersection area or the minimum distance is less than a preset threshold, a correlation edge is established between the corresponding instruction nodes. In the specific implementation process, the expected trajectory of each vehicle is represented as a four-dimensional space-time pipe, including three-dimensional space coordinates and one-dimensional time coordinates. The interference degree between instructions is quantified by calculating the overlapping volume and overlapping time of the space-time pipe. The higher the overlap degree, the greater the interference strength, and the greater the weight of the correlation edge. Another implementation way is to use a learning method based on graph neural network. A deep learning model is trained using historical scheduling data to automatically learn the complex nonlinear correlation relationship between instructions. This method can capture hidden interference patterns that traditional analysis methods cannot find, and improve the accuracy and intelligence level of the correlation graph construction.

[0067] S205, the time sequence rearrangement optimization unit performs time sequence rearrangement or spatial diversion on high-risk instructions; The time sequence rearrangement optimization unit rearranges the time sequence of high-risk instructions or spatially separates them, which means that for high-risk scheduling instructions identified in the space-time correlation graph, the conflict probability and aggregation risk between instructions are reduced by adjusting the execution time sequence of the instructions or modifying the driving path of the vehicle. The determination of high-risk instructions is based on the weight distribution and network topology characteristics in the correlation graph, and instructions corresponding to nodes with high degree centrality and large edge weight are usually selected as optimization objects. Time sequence rearrangement refers to adjusting the start execution time, path switching time point or arrival time of the instruction without changing the basic content of the instruction, so as to avoid space-time conflicts through time dimension coordination. Spatial separation refers to re-planning the driving path for the conflicting instructions, selecting alternative roads or adjusting the passing nodes, and reducing vehicle aggregation through spatial separation. The optimization process needs to consider multiple constraint conditions, including time window constraints, road capacity constraints, vehicle performance constraints, customer demand constraints, etc. The optimization objectives are usually to minimize the total transportation time, reduce the aggregation risk level, balance the road load, etc. Specific optimization algorithms can choose heuristic methods such as genetic algorithm, simulated annealing, particle swarm optimization, or use linear programming, integer programming, etc. Precise optimization methods, algorithm selection and parameter setting are adjusted according to problem size and time requirements, which are not limited here.

[0068] Preferably, the time sequence rearrangement optimization can use a local search algorithm based on tabu search, taking the current instruction time sequence arrangement as the initial solution, generating neighborhood solutions by swapping the execution order of adjacent instructions or adjusting the time parameters of a single instruction, evaluating the objective function value and constraint satisfaction of each neighborhood solution, and selecting the optimal feasible solution as the starting point for the next search. To avoid getting stuck in local optimum, the algorithm maintains a tabu table to record recently visited solutions to prevent repeated searches, and sets a desire criterion to allow accepting some solutions that are tabu but significantly improve the objective function. The spatial separation optimization finds the optimal alternative path for each conflicting vehicle through dynamic programming, considers factors such as path length, travel time, road congestion, etc. to construct a comprehensive cost function, and selects the path scheme with the minimum cost under the premise of meeting the time window and load constraints. Another implementation is to use a multi-objective optimization method, considering multiple objectives such as transportation efficiency, energy cost, aggregation risk, etc. to find a balanced solution set using the Pareto optimal theory, providing decision makers with multiple trade-off options.

[0069] S206, the feedback learning adaptation unit updates the space-time scene model parameters; The feedback learning adaptation unit updating the spatio-temporal scenario model parameters refers to self-adaptively adjusting and optimizing various model parameters in the digital twin scenario according to the actual execution effect optimized according to the dispatching instruction and the system operation feedback information, so as to improve the prediction accuracy and adaptability of the model. The model parameters include the speed-density relationship parameters in the traffic flow model, the acceleration and braking parameters in the vehicle motion model, the traffic capacity parameters in the road capacity model, the threshold parameters in the aggregation identification model, and the like. The feedback information sources include real-time monitoring data such as vehicle actual driving track, arrival time deviation, road congestion condition, and dispatching instruction execution success rate. The learning adaptation process adopts an online learning mode, can continuously receive new observation data and update the model parameters, and does not need to stop system operation for offline training. The parameter updating strategy needs to balance the learning speed and stability, avoid parameter oscillation caused by noise data, and ensure that the model can quickly adapt to environmental changes. The learning algorithm can select gradient descent, Bayesian inference, Kalman filtering, and the like, and the specific algorithm type and learning rate setting are determined according to the parameter characteristics and data characteristics. The adaptation process also needs to consider the update priority and influence range of different parameters, a more stringent verification mechanism is adopted for key parameters, and the learning period and update frequency can be dynamically adjusted according to the system load and data availability, which is not limited here.

[0070] Preferably, the model parameter updating can adopt a deep reinforcement learning method based on experience replay, an experience buffer is established to store historical state-action-reward triplets, a neural network is used to approximate the state value function or the action value function, and the network parameters are updated through time difference learning. In the specific implementation process, the current scenario state, the optimization decision and the execution result are encoded into a vector form and input into a deep neural network for value evaluation, the time difference error is calculated according to the actual reward signal, the network weights are updated using the back propagation algorithm, so as to realize the self-adaptive optimization of the model parameters. The experience replay mechanism trains the historical data by random sampling, breaks the time correlation between the data, and improves the learning efficiency and stability. Another implementation mode is to adopt an ensemble learning method, maintain multiple parallel model parameter versions, each version adopts different learning strategies or initialization modes, and integrate the prediction results of multiple models through a voting mechanism or a weighted average method. This integrated mode can reduce the bias and variance of a single model and improve the robustness of the overall prediction.

[0071] S207, generating a final dispatching control instruction based on the optimized dispatching instruction.

[0072] Based on the optimized scheduling instruction generation final scheduling control instruction is to convert the scheduling scheme after time sequence rearrangement, space diversion and parameter adaptation optimization into specific control command which can be directly issued to vehicle for execution. The final scheduling control instruction contains vehicle number, task identification, detailed path information, time node requirement, speed control parameter, checkpoint position, exception handling plan and other complete execution guidance information. The instruction generation process needs to perform format conversion and protocol adaptation to ensure that the instruction content meets the receiving format and communication protocol requirements of the vehicle terminal device. The control instruction also needs to contain execution priority, effective time range, cancellation condition and other management information to provide support for subsequent instruction adjustment and emergency handling. The integrity check of the instruction includes path connectivity verification, time logic consistency check, resource availability confirmation and other steps to ensure that the generated instruction is technically feasible and logically reasonable. The instruction issuing adopts the way of batch and priority, which gives priority to processing vehicles on emergency tasks and critical paths, while considering the bandwidth limitation of communication network and the receiving capacity of vehicle. The specific coding format, transmission protocol, encryption method and other technical details of the instruction are determined according to the system architecture and security requirements, which are not limited here.

[0073] Preferably, the scheduling control instruction generation can adopt the instruction synthesis method based on template, define the instruction template of different types of tasks in advance, contain the fixed format field and the placeholder of variable parameter, fill the variable parameter in the template according to the optimized scheduling scheme, and automatically generate the standardized control instruction. In the specific implementation process, the instruction template library is established to store the standard format of various task types, the appropriate template is selected according to the vehicle type, task nature and route characteristics, then the specific parameter value in the optimization scheme is mapped to the corresponding field of the template, finally the format verification and integrity check are performed to ensure that the generated instruction meets the execution requirements. The templating method can ensure the consistency and standardization of the instruction format, and reduce the risk of instruction parsing error. Another implementation way is to adopt the instruction generation technology based on semantics, use natural language processing method to convert the optimized scheduling scheme into instruction description close to natural language, and then convert it into machine executable control code through semantic parser. This way has better readability and flexibility, which is convenient for manual verification and debugging.

[0074] In the above embodiment, by adding the instruction timing simulation unit and the collaborative optimization unit in the intelligent scheduling module, the dynamic coupling effect analysis and optimization of multi-vehicle scheduling instructions are realized. First, a digital twin environment is established through the space-time scene construction unit, then the scheduling instructions are projected into the environment for simulation by the instruction execution simulation unit, and the potential vehicle aggregation risk is identified by the aggregation situation analysis unit, then the time-space correlation graph reflecting the mutual influence between instructions is constructed by the instruction interference evaluation unit, based on the correlation graph, the timing adjustment or path optimization of high-risk instructions is performed by the timing rearrangement optimization unit, finally the model parameter is continuously optimized by the feedback learning adaptation unit, thereby avoiding the traffic microcirculation problem in the multi-vehicle scheduling process, and improving the logistics transportation efficiency in the urban dense area.

[0075] The system in the embodiments of the present application will be described from the perspective of hardware processing. Please refer to Figure 10 The entity device structure diagram of a manufacturing contract logistics data fusion and scheduling system based on Beidou provided by the embodiments of the present application.

[0076] It should be noted that, Figure 10 The structure of the system shown is only an example, and should not bring any limitation to the function and use range of the embodiments of the present application.

[0077] As Figure 10 shown, the system includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 1002 or programs loaded from a storage portion 1008 into a random access memory (RAM) 1003, such as performing the methods in the above embodiments. In the RAM 1003, various programs and data required for system operation are also stored. The CPU 1001, the ROM 1002, and the RAM 1003 are connected to each other through a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0078] The following components are connected to the I / O interface 1005: an input section 1006 including a camera, a microphone, and the like; an output section 1007 including a liquid crystal display (LCD), a speaker, and the like; a storage section 1008 including a hard disk and the like; and a communication section 1009 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the I / O interface 1005 as necessary. A removable medium 1011 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 1010 as necessary, so that a computer program read out therefrom is installed in the storage section 1008 as necessary.

[0079] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing a computer program for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication section 1009, and / or installed from the removable medium 1011. When the computer program is executed by the central processing unit (CPU) 1001, various functions defined in the present application are executed.

[0080] It should be noted that the computer-readable medium in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer-readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries computer-readable computer programs. Such a propagated data signal can take many forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above.

[0081] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. Each block in the flowcharts or block diagrams can represent a module, a program segment, or a part of code containing one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders than that shown in the drawings. For example, two blocks that are shown in succession can actually be executed substantially in parallel, and sometimes in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams or flowcharts, and the combination of blocks in the block diagrams or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0082] As another aspect, the present application also provides a computer readable storage medium, which can be included in the system described in the above embodiments, or can exist independently without being assembled into the system. The above storage medium carries one or more computer programs, which, when executed by a processor of a system, enable the system to implement the method provided in the above embodiments.

[0083] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still make modifications to the technical solutions recorded in the foregoing embodiments, or make equivalent replacements to some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0084] In the above embodiments, according to the context, the term "when" can be interpreted as "if" or "after" or "in response to determining" or "in response to detecting". Similarly, according to the context, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)".

[0085] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (for example, floppy disk, hard disk, magnetic tape), optical media (for example, DVD), or semiconductor media (for example, solid state disk) and the like.

[0086] Those skilled in the art can understand that all or part of the processes in the above-mentioned method embodiments can be implemented by a computer program instructing relevant hardware to complete, the program can be stored in a computer readable storage medium, and the program can include the processes of the above-mentioned method embodiments when executed. The aforementioned storage medium includes ROM or random storage memory RAM, magnetic disc or optical disc and various storage code medium.

Claims

1. A BeiDou-based method for data fusion and scheduling of manufacturing contract logistics, applied to a system, characterized in that: The system includes a basic geographic and road network data module, a BeiDou dynamic monitoring data module, a cargo and contract data module, a logistics node data module, a data fusion processing module, and an intelligent scheduling module. The method includes: The basic geographic and road network data module acquires basic geographic data; the Beidou dynamic monitoring data module collects transport vehicle data; the cargo and contract data module stores cargo and contract data; and the logistics node data module acquires logistics node data. The data fusion processing module uses BeiDou positioning coordinates as a spatiotemporal reference to perform spatial matching on the basic geographic data, the transport vehicle data, the cargo and contract data, and the logistics node data to obtain a manufacturing logistics data association table. The data fusion processing module performs data fusion on the basic geographic data, the transport vehicle data, the cargo and contract data, and the logistics node data based on the manufacturing logistics data association table to obtain a unified dataset. The intelligent scheduling module generates the optimal scheduling control command based on the unified dataset and preset constraints, and sends the optimal scheduling control command to the corresponding vehicle terminal.

2. The method according to claim 1, characterized in that, The basic geographic and road network data module includes a geographic and road network information module, a format conversion module, a coordinate conversion module, a fusion and edge-matching processing module, and a data fusion module. The steps for acquiring basic geographic data by the basic geographic and road network data module specifically include: The geographic and road network information module acquires basic geographic information and road network information data; The format conversion module performs format unification processing on the basic geographic information and the road network information data to obtain unified format data; The coordinate transformation module uses the BeiDou high-precision control network as the spatial data basis to convert the unified format data into a preset geodetic coordinate system to obtain unified coordinate data. The fusion and edge processing module performs consistency processing on the geometric and attribute information of overlapping areas in the coordinate unified data to obtain consistent data. The data fusion module extracts transportation node data from the consistent data and performs fusion processing to obtain basic geographic data.

3. The method according to claim 1, characterized in that, The BeiDou dynamic monitoring data module includes a BeiDou dynamic monitoring information module, a first data preprocessing module, a first data clustering and fusion module, a first data association module, and a first data display module. The steps for the BeiDou dynamic monitoring data module to collect transport vehicle data specifically include: The Beidou dynamic monitoring information module collects raw vehicle data of the transport vehicle through the vehicle-mounted Beidou terminal; The first data preprocessing module performs noise reduction on the original vehicle data through nonlinear filtering to obtain preprocessed data; The first data clustering and fusion module performs spatiotemporal clustering on the preprocessed data to obtain vehicle trajectory data; The first data association module inputs the vehicle operation trajectory data into a preset frequent association model to obtain transport vehicle data, and the first data display module displays the transport vehicle data.

4. The method according to claim 1, characterized in that, The goods and contract data module includes a goods and contract information module, a second data preprocessing module, a data storage module, and a second data display module. The steps for storing goods and contract data in the goods and contract data module specifically include: The cargo and contract information module collects electronic waybill information and electronic contract documents to obtain raw business data. The cargo and contract information module extracts cargo information and contract information from the original business data; The second data preprocessing module standardizes the cargo information and the contract information to obtain cargo and contract data; The data storage module stores the goods and contract data, and the second data display module displays the goods and contract data.

5. The method according to claim 1, characterized in that, The logistics node data module includes a logistics node information module, a third data preprocessing module, a third data clustering and fusion module, a third data association module, and a third data display module. The steps for the logistics node data module to acquire logistics node data specifically include: The logistics node information module collects static node data and operational status information of the logistics nodes; The third data preprocessing module performs standardization processing on the node static data to obtain cleaned data. The standardization processing includes removing duplicate data, supplementing missing data, and unifying the data format. The third data clustering and fusion module classifies and aggregates the cleaning data based on geographical location and operational capability level to obtain clustered data. The third data association module associates the operation status information and the clustering data to obtain logistics node data, and the third data display module displays the logistics node data.

6. The method according to claim 1, characterized in that, The data fusion processing module includes a spatial matching module, a data attribute association module, a data configuration module, a data slicing deployment module, and a first service publishing module. Using BeiDou positioning coordinates as a spatiotemporal reference, the data fusion processing module performs spatial matching on the basic geographic data, the transport vehicle data, the cargo and contract data, and the logistics node data to obtain a manufacturing logistics data association table. The step of fusing the basic geographic data, transport vehicle data, cargo and contract data, and logistics node data based on the manufacturing logistics data association table to obtain a unified dataset specifically includes: The spatial matching module establishes a spatiotemporal reference system using the BeiDou positioning coordinates as a spatiotemporal reference. The spatial matching module maps the basic geographic data, the transport vehicle data, the cargo and contract data, and the logistics node data to the spatiotemporal reference system to obtain the mapped spatiotemporal reference system. The data attribute association module performs attribute association on the basic geographic data, the transport vehicle data, the cargo and contract data, and the logistics node data based on the mapped spatiotemporal reference system to obtain a manufacturing logistics data association table. The data configuration module merges data with relationships according to the manufacturing logistics data association table to obtain a fused dataset; The data slicing deployment module slices the fused dataset according to preset business requirements, and deploys the sliced ​​fused dataset to the corresponding business server to obtain a unified dataset, which is then published by the first service publishing module.

7. The method according to claim 1, characterized in that, The intelligent scheduling module includes an information fusion module, a scheduling matching module, a dynamic adjustment module, a scheme generation module, and a second service publishing module. The intelligent scheduling module generates an optimal scheduling control command based on the unified dataset and preset constraints, and sends the optimal scheduling control command to the corresponding vehicle terminal. Specifically, this includes the following steps: The fusion information module reads vehicle location information, cargo information, road network information, and node information from the unified dataset, and performs data preprocessing and format conversion to obtain basic scheduling data. The scheduling matching module performs multi-objective optimization calculations on the basic scheduling data according to preset constraints to obtain an initial scheduling scheme; The dynamic adjustment module optimizes the initial scheduling scheme based on the acquired dynamic data to obtain an optimized scheduling scheme. The scheme generation module converts the optimized scheduling scheme into a standard scheduling instruction format and generates the optimal scheduling control instruction. The second service publishing module distributes the optimal scheduling control command to the corresponding vehicle terminal.

8. A BeiDou-based manufacturing contract logistics data fusion and scheduling system, characterized in that, The system includes: One or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the system, the system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the system, the system performs the method as described in any one of claims 1-7.

Images