Business processing method of smart factory platform based on beidou grid code

Abstract

This invention provides a business processing method for a smart factory platform based on BeiDou grid codes. The method involves multi-level grid subdivision of the factory's physical space using BeiDou grid codes to generate unique grid codes. These grid codes represent location information in integer format. Equipment sensor data, personnel positioning data, and business logic data are bound to the grid codes to construct a spatiotemporal joint index table. Real-time correction of positioning data is performed using grid neighborhood topology relationships and personnel movement speed constraints to generate high-precision positioning trajectories. A rule engine maps business logic to grid codes, triggering automated operations based on grid attributes. By replacing traditional floating-point latitude and longitude coordinates with integer codes and optimizing data computation efficiency through bitwise operations, this method supports cross-level business associations and rapid path planning, thereby improving business response efficiency.

Description

Technical Field

[0001] This invention relates to the field of industrial digitization and digital twin technology, and in particular to a business processing method for a smart factory platform based on Beidou grid codes. Background Technology

[0002] With the gradual application of industrial digitalization and digital twin technologies in factory management, traditional smart factory systems have revealed several technical shortcomings in actual operation, making it difficult to meet the demands of modern factory management that requires high precision, low cost, and high collaboration. Specific problems include:

[0003] First, positioning accuracy relies excessively on hardware. Traditional systems require the deployment of dedicated positioning hardware such as UWB (Ultra-Wideband) and LoRa (LoRa) to achieve spatial positioning of personnel and equipment. This not only significantly increases initial deployment and maintenance costs but also results in unsatisfactory positioning performance even with substantial hardware investment. For example, when using Bluetooth positioning technology, the positioning error is typically 2-5 meters. In scenarios with high positioning accuracy requirements, such as chemical plants and warehouses, inaccurate positioning can easily lead to safety and efficiency issues, such as personnel accidentally entering dangerous areas or equipment scheduling deviations. Second, data fragmentation is severe. In traditional systems, equipment status data, personnel trajectory data, and business logic data belong to different data systems, lacking a unified spatial benchmark for correlation and integration. This prevents data from being linked and analyzed. For example, when equipment sensors detect excessive gas concentration in a certain area, the system cannot quickly match the location of personnel and evacuation routes within that area. Manual intervention is required to query multiple types of data before a response plan can be formulated, significantly reducing business response efficiency.

[0004] Finally, the functional limitations are obvious. Traditional digital twin systems mostly remain at the level of three-dimensional visualization, only able to display the physical spatial layout of the factory and the appearance of equipment through models. They cannot drive real-time business responses. For example, although the system can visualize dangerous areas, it cannot automatically trigger safety warnings based on personnel location data; work order scheduling requires manual allocation of tasks according to production progress, and cannot achieve dynamic dispatching by combining equipment location and personnel trajectory; path planning can only generate routes based on fixed maps and cannot avoid temporary obstacles in real time. As a result, digital twin technology has failed to fully support factory operations and cannot meet the needs of modern factories for automated decision-making and real-time business processing.

[0005] These shortcomings collectively restrict the improvement of the intelligence level of smart factories, and there is an urgent need for a new business processing method that can break through hardware dependence, achieve data integration, and expand business functions. Summary of the Invention

[0006] The main objective of this invention is to provide a business processing method for a smart factory platform based on BeiDou grid codes, which solves the problem that traditional smart factory systems have exposed many technical defects in actual operation and are difficult to meet the needs of modern factory management with high precision, low cost and high collaboration.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a business processing method for a smart factory platform based on Beidou grid codes, the method comprising:

[0008] S1. The factory's physical space is divided into multi-level grids using BeiDou grid codes to generate unique grid codes; the grid codes use integer format to represent location information.

[0009] S2. Bind equipment sensor data, personnel positioning data, and business logic data with grid coding to construct a spatiotemporal joint index table;

[0010] S3. Real-time correction of positioning data is performed using grid neighborhood topology and personnel movement speed constraints to generate high-precision positioning trajectories;

[0011] S4. Map business logic to grid codes through the rules engine, and trigger automated operations based on grid attributes.

[0012] In the preferred scheme, step S1 involves multi-level gridding of the factory's physical space using BeiDou grid codes, including:

[0013] S11. Based on the functions of the factory area, an octree partitioning algorithm is used to generate multi-level grid cells; each grid cell corresponds to a different side length; a grid partitioning accuracy optimization algorithm is introduced to calculate the matching degree between the grid cell side length and the accuracy required by the regional business needs. The formula is:

[0014] ;

[0015] in For precision weighting system, To achieve the minimum positioning accuracy required for regional business needs. The side length of the grid cell. To ensure that a single grid cell effectively covers the business area, For the total area of ​​the region, when At that time, it is determined that the mesh subdivision accuracy meets the business requirements;

[0016] S12. Assign a unique code to each grid cell; the code includes the level identifier, longitude offset, and latitude offset;

[0017] S13. Dynamically adjust the grid level according to the regional function: use a high-resolution grid for passage areas, a medium-resolution grid for storage areas, and a low-resolution grid for open areas.

[0018] S14. Generate an occupied grid list, recording the grid cells occupied by buildings and obstacles; use a grid occupancy calculation algorithm to determine the probability that a grid is occupied. The formula is:

[0019] ;

[0020] in The number of times the sensor detects obstacles within the grid. For sensor detection weights, Mark the number of times obstacles exist within the grid on the factory map. Label the map with weights. For the total number of detections and annotations, when When that happens, add the grid to the list of occupied grids.

[0021] In the preferred embodiment, step S2 involves binding device sensor data, personnel positioning data, and business logic data with grid coding, including:

[0022] S21. Acquire equipment sensor data, including environmental parameters and equipment status; calculate data reliability using a sensor data reliability assessment algorithm. The formula is:

[0023] ;

[0024] in For a single sensor data acquisition, The average value of n collected data points. For the number of data collections, The maximum allowable deviation value for the sensor, when When the data is deemed reliable, it is used for subsequent binding operations;

[0025] S22. Obtain personnel positioning data, including coordinate information of continuous positioning points;

[0026] S23. Associate business logic data with grid coding to form a rule set;

[0027] S24. Construct a spatiotemporal joint index table to store the mapping relationship between grid coding and sensor data, positioning data and business logic data;

[0028] S25. Enables rapid retrieval and dynamic updating of multi-source data through index tables;

[0029] The retrieval response time of the index table is calculated using a retrieval efficiency optimization algorithm. The formula is:

[0030] ;

[0031] in For the basic retrieval time of the index table, For data correlation coefficient, This refers to the amount of data retrieved in a single search. For data transmission bandwidth, This formula optimizes the index table structure to improve the index table query hit rate. .

[0032] In the preferred embodiment, step S24 includes:

[0033] S241. For each grid code, associate the corresponding sensor data and positioning data;

[0034] S242. Generate a rule set based on business logic data, including security rules and scheduling rules; calculate the rule validity using a rule set validity evaluation algorithm. The formula is:

[0035] ;

[0036] in The number of times the rule is triggered. To trigger frequency weights, The number of times the rule is correctly executed. To implement correctness weights, For the total number of rules, when When the rule set is valid, it is determined that the rule set is valid.

[0037] S243. The mapping relationship between grid codes and rule sets is stored in a database;

[0038] S244. When sensor data is updated, the mapping relationship in the index table is dynamically adjusted; the sensitivity adjustment algorithm is used to calculate the adjusted sensitivity. The formula is:

[0039] ;

[0040] in Adjust the number of times for the mapping relationship. For the number of times sensor data is updated, To adjust the coefficient, when At that time, the determination of the mapping relationship adjustment is sensitive;

[0041] S245. Support cross-level business association queries through index tables.

[0042] In the preferred embodiment, step S3 includes:

[0043] S31. Obtain the grid code of continuous personnel positioning points and calculate the number of grid spans between adjacent positioning points;

[0044] S32. If the number of crossings exceeds the preset speed constraint threshold, it is determined to be positioning drift; a dynamic speed threshold calculation algorithm is introduced, the formula is:

[0045] ;

[0046] in As the speed constraint threshold, This represents the side length of the current level grid. This is the slope influence coefficient. The slope value of the area where the location point is located. This represents the maximum slope value within the factory area. This refers to the time interval for location data collection.

[0047] S33. Filter out illegal location points by combining the occupied grid list;

[0048] S34. Smooth the positioning data using Kalman filtering; introduce a filter gain optimization algorithm to calculate the Kalman filter gain. The formula is:

[0049] ;

[0050] in Let the prior error covariance be at time k. For the observation matrix, To observe the noise covariance, For speed influence coefficient, Let k be the speed at which people move. This refers to the maximum permissible movement speed of personnel within the factory.

[0051] S35. Converge the drifting positioning points to the nearest valid grid to generate a high-precision positioning trajectory; use a convergence distance calculation algorithm to determine the optimal convergence distance from the drifting points to the valid grid. The formula is:

[0052] ;

[0053] in Here are the coordinates of the drift point. The coordinates of the center point of the valid grid. Let be the convergence coefficient, when At that time, the drift point convergence is complete.

[0054] In the preferred embodiment, step S4 includes:

[0055] S41. Obtain grid attributes, including hazard attributes and equipment distribution information, through the rule engine; calculate attribute importance using a grid attribute importance evaluation algorithm. The formula is:

[0056] ;

[0057] in For the weight of dangerous attributes, The danger level is... For the weights of device distribution attributes, The importance level of the equipment;

[0058] S42. Determine whether the grid properties meet the preset conditions based on the sensor data;

[0059] S43. If the preset conditions are met, a safety warning or work order dispatch will be triggered;

[0060] S44. Generate dynamic path planning based on grid encoding and grid properties;

[0061] S45. Push route instructions to personnel's mobile terminals; evaluate the push effect using an instruction push success rate calculation algorithm. The formula is:

[0062] ;

[0063] in To indicate the number of successful push notifications. This represents the total number of push notifications. The response time impact coefficient, For terminal response time, For the maximum allowable response time, when At that time, the push notification effect was judged to be good.

[0064] In the preferred scheme, the dynamic path planning generated based on grid codes and grid attributes includes:

[0065] S411. Calculate the grid level distance between the starting point and the ending point using integer bitwise operations;

[0066] S412. The shortest path is generated using the A search algorithm; the heuristic function of the A algorithm is introduced to optimize the algorithm, and the heuristic function is:

[0067] ;

[0068] in, It is the total cost of node n. It is the actual cost from the starting point to node n. It is a heuristic cost estimate from node n to the target node;

[0069] The heuristic cost estimate is calculated using a chosen heuristic function, specifically the Manhattan distance, as shown in the formula:

[0070] ;

[0071] in, The x-axis coordinate of the center point of the current grid node n; The x-axis coordinate of the center point of the grid corresponding to the target grid node; The y-axis coordinate of the center point of the current grid node n; The y-axis coordinate of the center point of the target mesh node;

[0072] The optimized heuristic function is:

[0073] ;

[0074] in For heuristic weighting coefficients, The original Manhattan distance heuristic function, This represents the impact coefficient of grid cost. The cost value of the grid where node n is located. This represents the maximum grid cost value.

[0075] S413. Dynamically adjust path costs based on grid attributes, setting high costs for grids in restricted areas and decreasing costs for grids in risk areas based on distance from the center; calculate the cost of risk area grids using a risk area cost refinement algorithm. The formula is:

[0076] ;

[0077] in The greatest cost to the risk zone The distance from the risk zone grid to the risk center. This serves as a baseline distance for risk impact.

[0078] S414. Set negative costs for critical equipment meshes and prioritize planning through critical equipment meshes;

[0079] S415. Manage grid nodes using open and closed lists to determine the optimal path; calculate management efficiency using a node management efficiency evaluation algorithm. The formula is ,in This represents the number of nodes that have been processed. For the number of nodes in the open list, The number of nodes in the closed list. Based on the processing time, This refers to the actual processing time, when At that time, the node management efficiency is determined to be up to standard.

[0080] In the preferred scheme, after generating a unique grid code in S1, a grid code fault tolerance verification algorithm is introduced to verify the integrity and correctness of the grid code, including:

[0081] S15. Extract the hierarchy identifier from the mesh code. Longitude offset Latitude offset ;

[0082] S16. Calculate the code check value The formula is ,in The total number of bits in the code. To verify the modulus value;

[0083] S17. The calculated result The code is compared with the built-in check bit. If they match, the code is considered correct. If they do not match, the grid code is regenerated and steps S15-S17 are repeated until the code verification passes, ensuring that the grid code is transmitted without errors in subsequent data association and business processing.

[0084] In the preferred scheme, after generating a high-precision positioning trajectory in S3, a trajectory similarity analysis algorithm is introduced to verify the consistency of the positioning trajectory within a continuous time period, including:

[0085] S36. Select two adjacent time periods and Within the positioning trajectory, extract the grid-coded sequence on the trajectory respectively. and ,in The number of location points within each time period;

[0086] S37. Calculate the trajectory similarity Sim using the following formula:

[0087] ;

[0088] in The number of identical grid codes in the two sequences. for Length, for The length; and simultaneously introducing the trajectory fluctuation coefficient. The formula is:

[0089] ;

[0090] in for The Middle The center point coordinates of each grid code for The Middle The center point coordinates of each grid code This represents the side length of the current level grid.

[0091] S38. When and If the trajectory is consistent and the positioning data is stable, then step S3 is executed again to correct the positioning and further improve the reliability of the positioning trajectory.

[0092] In the preferred embodiment, after the automated operation is triggered by S4, an operation effect feedback evaluation algorithm is introduced to quantitatively evaluate the execution effect of the automated operation, including:

[0093] S46. Determine the type of automated operation and extract relevant parameters; for safety warning operations, extract the warning response time. Early warning accuracy Work order dispatch operation extracts work order completion rate. Work order completion time ; Path planning operation extracts path deviation rate Path time ;

[0094] S47. Calculate the overall evaluation value of the operation. The formula is:

[0095] ;

[0096] in To determine the number of parameters related to the operation, For the first The weights of each parameter,

[0097] in, Determined based on the type of operation, including safety warning operations. , ;

[0098] Work order dispatching in progress , ;

[0099] In the path planning operation , ;

[0100] For the first The standardized values ​​of each parameter For time-related parameters, smaller The larger the value, the better for parameters such as accuracy and completion rate. , For the first The maximum value of each parameter. For the first The minimum value of each parameter;

[0101] S48. When When the time is right, the automated operation is deemed to be working well; if Then, analyze the reasons for parameter deviations, adjust the business logic mapping relationship in the rule engine, repeat the S4 steps to optimize automated operations, and ensure that the business processing effect meets the factory management requirements.

[0102] This invention provides a business processing method for a smart factory platform based on BeiDou grid codes. This method effectively solves the problems of hardware-dependent positioning accuracy, data fragmentation, and functional limitations in traditional smart factory systems, and has many significant advantages:

[0103] Firstly, it significantly reduces deployment and maintenance costs. This method does not require additional deployment of dedicated positioning hardware such as UWB and LoRa. It only uses multi-level grid subdivision and positioning correction technology of Beidou grid code to improve positioning accuracy to sub-meter level without increasing hardware investment. It also uses grid neighborhood topology, personnel movement speed constraints and Kalman filter smoothing to reduce the initial hardware procurement and installation costs of the project and avoid the maintenance expenses of dedicated hardware in the later stage, thus significantly reducing the construction and operation costs of smart factories.

[0104] Secondly, it enables efficient fusion and linkage of multi-source data. By binding equipment sensor data, personnel positioning data, and business logic data with a unique grid code, a spatiotemporal joint index table is constructed. This breaks the fragmentation of data belonging to different systems in traditional systems, supports efficient retrieval, dynamic updates, and cross-level business association queries of multi-source data. When an anomaly occurs in a certain grid area, the system can quickly associate the personnel location, equipment status, and preset business rules within that grid, providing a unified and accurate spatiotemporal data benchmark for subsequent business processing and significantly improving the efficiency of data linkage analysis.

[0105] Third, this method significantly expands the business functions of the digital twin system and enhances its automated decision-making capabilities. It maps business rules to grid coding via a rule engine, upgrading the digital twin system from a traditional 3D visualization level to a platform with real-time business response capabilities. When grid attributes meet preset conditions, it can automatically trigger operations such as safety warnings and work order dispatch. In path planning scenarios, based on integer bitwise operations and the A-search algorithm using grid coding, combined with grid attributes to dynamically adjust path costs, it can generate the optimal path in real time and push it to personnel's mobile terminals. This avoids temporary obstacles and completes dynamic path planning without manual intervention, effectively achieving automated decision-making and fully leveraging the supporting role of digital twin technology for factory operations. Simultaneously, by using BeiDou grids to carry time data to form a BeiDou spatiotemporal grid, it is possible to trace changes in object attributes within the grid, providing a spatiotemporal dimension basis for judging the safety of business operations and further ensuring factory operational safety.

[0106] Furthermore, the mesh partitioning in this method can dynamically adjust the hierarchy according to the functions of different factory areas, adapting to the business needs and accuracy requirements of different areas. In addition, the data calculation uses integer bitwise operations instead of traditional floating-point latitude and longitude operations, optimizing the data calculation efficiency and ensuring faster and more accurate business response. It is suitable for the intelligent management of smart factories in multiple fields such as chemical industry, warehousing, and manufacturing, and can significantly improve the factory's operational efficiency and safety management level. Attached Figure Description

[0107] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0108] Figure 1 This is a schematic diagram illustrating the correspondence between different levels of BeiDou grid codes within a factory, as presented in the smart factory platform based on BeiDou grid codes of this invention.

[0109] Figure 2 This is a schematic diagram of the grid marked with obstacle areas in the smart factory platform based on Beidou grid codes of this invention;

[0110] Figure 3 This is a schematic diagram of the interface of the smart factory platform based on Beidou grid code of this invention in the digital twin scenario of Xingrui Silicon Materials. Detailed Implementation

[0111] Example 1

[0112] like Figure 1-3 As shown, a business processing method for a smart factory platform based on BeiDou grid codes is described, the method including:

[0113] S1. The factory's physical space is divided into multi-level grids using BeiDou grid codes to generate unique grid codes; the grid codes use integer format to represent location information.

[0114] S2. Bind equipment sensor data, personnel positioning data, and business logic data with grid coding to construct a spatiotemporal joint index table;

[0115] S3. Real-time correction of positioning data is performed using grid neighborhood topology and personnel movement speed constraints to generate high-precision positioning trajectories;

[0116] S4. Map business logic to grid codes through the rules engine, and trigger automated operations based on grid attributes.

[0117] The usage method of the smart factory platform business processing method based on Beidou grid code is as follows: When applying this method, first execute step S1. Based on the factory building structure and the business needs of different areas, use the octree partitioning algorithm to perform multi-level grid partitioning of the factory physical space. Assign a unique integer format code to each grid. The code includes the level identifier, longitude offset and latitude offset. In this way, the factory space is transformed into a grid system that can be located through integer codes. For example, the passage area uses an eight-level grid of 0.97m×0.97m and the open area uses a six-level grid of 61.84m×61.84m.

[0118] Next, step S2 is performed to collect environmental parameters and equipment status data collected by sensors in the factory, continuous coordinate data transmitted from personnel positioning terminals, and business logic data such as safety warning rules and work order scheduling tasks. These multi-source data are bound one by one to the grid codes generated in S1 to build a spatiotemporal joint index table. The index table enables rapid data retrieval and dynamic updates. For example, data from a gas sensor can be directly associated with the code of its grid for easy subsequent retrieval.

[0119] Then, step S3 is carried out. Based on the neighborhood topology relationship of the grid in S1 and the personnel movement speed constraint, the real-time collected personnel positioning data is processed to calculate the number of grid crossings between adjacent positioning points. If the number of crossings exceeds the speed constraint threshold, it is determined to be positioning drift. Combining the list of grids occupied by buildings and obstacles, illegal positioning points are converged to the nearest legal grid. At the same time, Kalman filtering is used to smooth the data, and finally a high-precision positioning trajectory is generated.

[0120] Finally, the S4 step is implemented, using the rule engine to map the preset business logic to the corresponding grid code. When the grid attribute meets the trigger condition, the operation is automatically executed. For example, if a grid is marked as a high-risk area and personnel are detected entering, the system will automatically trigger a security warning. During path planning, the distance is calculated based on the integer operation of the grid code, and the optimal path is generated by combining the A search algorithm with the grid cost and pushed to the personnel terminal.

[0121] The beneficial effects of this method are reflected in several aspects: First, it significantly reduces costs. It eliminates the need for additional deployment of dedicated positioning hardware such as UWB and LoRa. Positioning accuracy can be improved to sub-meter level simply through the segmentation and correction logic of the BeiDou grid code, reducing capital investment in hardware procurement and maintenance. Simultaneously, the integer-formatted grid code replaces traditional floating-point latitude and longitude coordinates, optimizing data computation efficiency through bitwise operations and reducing system computing resource consumption. Second, it solves the problem of data silos. Through the spatiotemporal joint index table constructed in step S2, equipment, personnel, and business data are deeply bound to the grid code, achieving multi-source data linkage and integration. This avoids the coordination difficulties caused by data belonging to different systems in traditional systems. For example, when a sensor in a grid detects a gas anomaly, it can quickly associate the location of personnel within that grid with preset safety rules, improving data application efficiency. Third, it enhances business responsiveness and automation. The combination of rule engine and grid coding in the S4 step upgrades the digital twin system from simple visualization to a platform with real-time decision-making capabilities. It can automatically trigger operations such as early warning, order dispatch, and path planning without manual intervention. For example, when a temporary obstacle appears, the system can quickly mark the corresponding grid as a prohibited state and replan the path, improving the flexibility and safety of factory operations. It is suitable for various scenarios with high requirements for space management and business responsiveness, such as chemical, warehousing, and manufacturing industries, and helps factories achieve intelligent and efficient management.

[0122] In the preferred scheme, step S1 involves multi-level gridding of the factory's physical space using BeiDou grid codes, including:

[0123] S11. Based on the functions of the factory area, an octree partitioning algorithm is used to generate multi-level grid cells; each grid cell corresponds to a different side length; a grid partitioning accuracy optimization algorithm is introduced to calculate the matching degree between the grid cell side length and the accuracy required by the regional business needs. The formula is:

[0124] ;

[0125] in For precision weighting system, To achieve the minimum positioning accuracy required for regional business needs. The side length of the grid cell. To ensure that a single grid cell effectively covers the business area, For the total area of ​​the region, when At that time, it is determined that the mesh subdivision accuracy meets the business requirements;

[0126] S12. Assign a unique code to each grid cell; the code includes the level identifier, longitude offset, and latitude offset;

[0127] S13. Dynamically adjust the grid level according to the regional function: use a high-resolution grid for passage areas, a medium-resolution grid for storage areas, and a low-resolution grid for open areas.

[0128] S14. Generate an occupied grid list, recording the grid cells occupied by buildings and obstacles; use a grid occupancy calculation algorithm to determine the probability that a grid is occupied. The formula is:

[0129] ;

[0130] in The number of times the sensor detects obstacles within the grid. For sensor detection weights, Mark the number of times obstacles exist within the grid on the factory map. Label the map with weights. For the total number of detections and annotations, when When that happens, add the grid to the list of occupied grids.

[0131] Specific implementation steps:

[0132] The overall process revolves around grid cell generation, encoding allocation, hierarchical adjustment, and grid occupancy recording. A dedicated algorithm is introduced to optimize partitioning accuracy and occupancy determination accuracy. Specifically, step S11 is executed first. Based on the functional attributes of different areas of the factory, an octree partitioning algorithm is used to decompose the factory's physical space into multiple levels of grid cells, with each level of grid cell corresponding to a fixed side length (0.97m for level 8 grid and 61.84m for level 6 grid). To ensure that the grid cell side length matches the accuracy required by regional business needs, a grid partitioning accuracy optimization algorithm is introduced, which calculates the matching degree... The formula for determining whether the partitioning meets the standards is as follows: ,in This is the accuracy weighting coefficient, with a value ranging from 0.6 to 0.8, used to balance the weight between accuracy requirements and coverage efficiency. The minimum positioning accuracy required for regional business needs, expressed in meters, represents the minimum positional accuracy standard required for conducting business in this region. This represents the side length of the mesh cell, in meters (m), which is the actual side length of the mesh generated by the current subdivision. The effective coverage area of ​​a single grid cell is the service area, in units of This refers to the area within a single grid that can be used to carry out the target business. The total area of ​​the region is expressed in units of 1000 sq. That is, the total area of ​​the factory area currently being divided; when the calculated... If the current grid subdivision accuracy is sufficient to meet the business requirements of the area, then the grid cell side lengths need to be adjusted and the grid subdivision re-done.

[0133] Next, step S12 is executed, assigning a unique code to each grid cell generated in S11. This code contains three core pieces of information: a level identifier (used to distinguish the subdivision level to which the grid belongs, such as level 8 or level 6), a longitude offset (reflecting the distance the grid is offset from the reference point in the longitude direction), and a latitude offset (reflecting the distance the grid is offset from the reference point in the latitude direction). This code allows for the precise location of each grid cell in the factory space.

[0134] Then, step S13 is executed to dynamically adjust the grid level based on the functional differences of different areas in the factory. For passageways, where frequent personnel and equipment movement requires high-precision positioning, a high-resolution grid (level 8 grid) is used. For storage areas, which need to balance positioning accuracy and coverage, a medium-resolution grid (level 7 grid) is used. For open areas with lower positioning accuracy requirements, a low-resolution grid (level 6 grid) is used, thus achieving a reasonable allocation of grid resources. Finally, step S14 is executed to generate an occupied grid list, which records the grid cells occupied by buildings and obstacles within the factory. To avoid misjudgments, a grid occupancy calculation algorithm is used to determine the probability of a grid being occupied. Its calculation formula is ,in This refers to the number of times the sensors detect obstacles within the grid, i.e., the number of positive detections of whether there are obstacles in the grid by various sensors (infrared sensors, lidar) deployed in the factory. The sensor detection weight, with a value range of 0.7-0.9, reflects the reliability weight of the sensor detection results; Mark the number of times obstacles exist within the grid on the factory map, that is, mark the number of times obstacles exist in the grid based on factory design drawings, on-site surveys, etc. The map label weights range from 0.3 to 0.5, reflecting the credibility weight of the map label information. The total number of tests and annotations, i.e. and The sum;

[0135] When the calculation yields If the grid cell is indeed occupied by a building or obstacle, it will be included in the list of occupied grid cells, and these grid cells will be avoided in subsequent positioning, path planning and other operations.

[0136] In the preferred embodiment, step S2 involves binding device sensor data, personnel positioning data, and business logic data with grid coding, including:

[0137] S21. Acquire equipment sensor data, including environmental parameters and equipment status; calculate data reliability using a sensor data reliability assessment algorithm. The formula is:

[0138] ;

[0139] in For a single sensor data acquisition, The average value of n collected data points. For the number of data collections, The maximum allowable deviation value for the sensor, when When the data is deemed reliable, it is used for subsequent binding operations;

[0140] S22. Obtain personnel positioning data, including coordinate information of continuous positioning points;

[0141] S23. Associate business logic data with grid coding to form a rule set;

[0142] S24. Construct a spatiotemporal joint index table to store the mapping relationship between grid coding and sensor data, positioning data and business logic data;

[0143] S25. Enables rapid retrieval and dynamic updating of multi-source data through index tables;

[0144] The retrieval response time of the index table is calculated using a retrieval efficiency optimization algorithm. The formula is:

[0145] ;

[0146] in For the basic retrieval time of the index table, For data correlation coefficient, This refers to the amount of data retrieved in a single search. For data transmission bandwidth, This formula optimizes the index table structure to improve the index table query hit rate. .

[0147] In the preferred embodiment, step S24 includes:

[0148] S241. For each grid code, associate the corresponding sensor data and positioning data;

[0149] S242. Generate a rule set based on business logic data, including security rules and scheduling rules; calculate the rule validity using a rule set validity evaluation algorithm. The formula is:

[0150] ;

[0151] in The number of times the rule is triggered. To trigger frequency weights, The number of times the rule is correctly executed. To implement correctness weights, For the total number of rules, when When the rule set is valid, it is determined that the rule set is valid.

[0152] S243. The mapping relationship between grid codes and rule sets is stored in a database;

[0153] S244. When sensor data is updated, the mapping relationship in the index table is dynamically adjusted; the sensitivity adjustment algorithm is used to calculate the adjusted sensitivity. The formula is:

[0154] ;

[0155] in Adjust the number of times for the mapping relationship. For the number of times sensor data is updated, To adjust the coefficient, when At that time, the determination of the mapping relationship adjustment is sensitive;

[0156] S245. Support cross-level business association queries through index tables.

[0157] The specific implementation steps revolve around the acquisition of multi-source data, credibility verification, association binding, index construction, and optimization. The details are as follows: First, the operation of "binding equipment sensor data, personnel positioning data, and business logic data with grid coding" is performed, with steps S21 to S25 in sequence: Step S21 involves acquiring equipment sensor data, which includes environmental parameters and equipment status within the factory; to ensure data reliability, a sensor data credibility assessment algorithm is used to calculate data credibility. The calculation formula is: ,in This represents the data value collected by a single sensor scan. This represents the average value of the data collected n times for this parameter. Represents the number of data collections and must meet the following requirements: , This represents the maximum permissible deviation value specified by the sensor at the factory; when calculated... At this point, the sensor data is determined to be reliable and can be used for subsequent binding operations with grid codes. Step S22 involves acquiring personnel positioning data, which includes the coordinate information of continuous positioning points transmitted back by personnel-carried positioning terminals, providing basic data for subsequent binding with grid codes and positioning correction. Step S23 involves associating business logic data with grid codes to form a rule set. The business logic data includes safety warning conditions, work order scheduling rules, etc. By associating with grid codes, the business rules can be accurately mapped to specific spatial locations within the factory.

[0158] Step S24 involves constructing a spatiotemporal composite index table. The core function of this index table is to store the mapping relationships between grid codes and sensor data, positioning data, and business logic data, thus enabling the association of different types of data with spatial locations. Step S25 utilizes this spatiotemporal composite index table to achieve rapid retrieval and dynamic updates of multi-source data. To improve retrieval efficiency, a retrieval efficiency optimization algorithm is needed to calculate the index table's retrieval response time. The calculation formula is: ,in This represents the basic retrieval time of the index table, in milliseconds (ms), which is the basic query time without considering factors such as data volume and bandwidth. The data correlation coefficient, ranging from 1.2 to 1.5, is used to correct the impact of multi-source data association on retrieval time. This represents the amount of data retrieved in a single search, measured in KB. This represents data transmission bandwidth, measured in KB / ms, which is the amount of data that can be transmitted per unit of time. This represents the index table query hit rate, ranging from 0.7 to 1.0, which is the proportion of index records that can be directly matched by a search request. The calculated hit rate can be improved by adjusting the index table structure. This meets the factory's real-time data retrieval requirements.

[0159] The step of "constructing a spatiotemporal joint index table to store the mapping relationship between grid codes and sensor data, location data, and business logic data" includes detailed steps S241 to S245: Step S241 involves associating each grid code with the corresponding sensor data and location data within that grid area, ensuring that each spatial location's grid code corresponds to specific monitoring and personnel data. Step S242 generates a rule set based on the business logic data, which includes security rules and scheduling rules; to verify the effectiveness of the rule set, a rule set effectiveness evaluation algorithm is used to calculate the rule effectiveness. The calculation formula is: ,in This represents the number of times the rule is triggered, that is, the number of times the rule meets the triggering condition; This represents the trigger frequency weight, with a value ranging from 0.4 to 0.6, reflecting the impact of rule trigger frequency on effectiveness; This represents the number of times the rule was executed correctly, that is, the number of times the business processing was completed as expected after the rule was triggered; This represents the weight of correct execution, with a value ranging from 0.6 to 0.8, reflecting the core impact of rule execution accuracy on effectiveness; This represents the total number of rules in the rule set; when calculated... When the rule set is deemed valid, it can be used for subsequent business processing.

[0160] Step S243 ensures persistent data storage and ease of retrieval and maintenance by storing the mapping relationship between grid codes and rule sets in the database.

[0161] Step S244 involves dynamically adjusting the mapping relationship between the corresponding grid code in the index table and the sensor data when the sensor data is updated. To ensure timely adjustment, a mapping relationship adjustment sensitivity calculation algorithm is used to calculate the adjustment sensitivity. The calculation formula is: ,in This represents the number of times the mapping relationship has been adjusted, that is, the number of times the mapping relationship corresponding to the index table has been modified after the sensor data is updated; This represents the number of times the sensor data has been updated; This represents an adjustment factor, ranging from 0.8 to 1.2, used to correct for delays or redundancy during the adjustment process; when the calculated... At the same time, the determination of the mapping relationship is sensitive and can respond promptly to changes in sensor data.

[0162] Step S245 enables cross-level business association queries through the spatiotemporal composite index table. This means that not only can various types of data corresponding to a single level grid be queried, but business data between different level grids can also be queried, meeting the multi-dimensional data query needs of complex business scenarios in factories.

[0163] In the preferred embodiment, step S3 includes:

[0164] S31. Obtain the grid code of continuous personnel positioning points and calculate the number of grid spans between adjacent positioning points;

[0165] S32. If the number of crossings exceeds the preset speed constraint threshold, it is determined to be positioning drift; a dynamic speed threshold calculation algorithm is introduced, the formula is:

[0166] ;

[0167] in As the speed constraint threshold, This represents the side length of the current level grid. This is the slope influence coefficient. The slope value of the area where the location point is located. This represents the maximum slope value within the factory area. This refers to the time interval for location data collection.

[0168] S33. Filter out illegal location points by combining the occupied grid list;

[0169] S34. Smooth the positioning data using Kalman filtering; introduce a filter gain optimization algorithm to calculate the Kalman filter gain. The formula is:

[0170] ;

[0171] in Let the prior error covariance be at time k. For the observation matrix, To observe the noise covariance, For speed influence coefficient, Let k be the speed at which people move. This refers to the maximum permissible movement speed of personnel within the factory.

[0172] S35. Converge the drifting positioning points to the nearest valid grid to generate a high-precision positioning trajectory; use a convergence distance calculation algorithm to determine the optimal convergence distance from the drifting points to the valid grid. The formula is:

[0173] ;

[0174] in Here are the coordinates of the drift point. The coordinates of the center point of the valid grid. Let be the convergence coefficient, when At that time, the drift point convergence is complete.

[0175] The specific implementation steps are as follows: First, execute step S31 to obtain the grid code corresponding to the continuous positioning points transmitted back by the positioning terminal carried by the personnel. Based on the layer and coordinate offset information of the grid code, calculate the number of grids crossed between two adjacent positioning points to reflect the spatial position change of the personnel per unit time.

[0176] Next, step S32 is executed to determine whether the number of grid spans between adjacent positioning points exceeds a preset velocity constraint threshold. If it does, it is determined to be positioning drift. To make the velocity constraint threshold more closely match the actual terrain of the factory, a dynamic velocity threshold calculation algorithm is introduced. The calculation formula is as follows: ,in The speed constraint threshold is expressed in grids per second, representing the maximum number of grids that a person can reasonably move across per unit of time. This represents the current level of grid side length, in meters (m), which is the actual side length of the grid cell currently used for positioning. The slope influence coefficient, with a value range of 0.1-0.3, is used to correct the impact of terrain slope on the speed of personnel movement. This represents the slope value of the area where the location point is located, in degrees, reflecting the degree of inclination of the terrain in that area.

[0177] This represents the maximum slope value within the factory area, expressed in degrees; that is, the maximum slope value across all areas of the factory. The location data acquisition time interval is in seconds, representing the time difference between two location data acquisitions. Then, step S33 is executed, combining the previously generated occupied grid list to record the grid cells occupied by buildings and obstacles. Illegal location points in the location data are filtered out; if the grid cell corresponding to a location point is in the occupied grid list, the location point is determined to be illegal and will not participate in subsequent trajectory generation, thus preventing the location of personnel inside buildings or obstacles.

[0178] Next, step S34 is executed, where the filtered positioning data is smoothed using a Kalman filter to reduce random errors. To optimize the filtering effect, a filter gain optimization algorithm is introduced to calculate the Kalman filter gain. The calculation formula is: ;

[0179] in Let be the prior error covariance at time k, representing the error variance in estimating the state at time k based on the state at time k-1; This is the observation matrix, used to map the state vector to the observation vector; This is the transpose of the observation matrix; Let k be the observation noise covariance, representing the noise level of the data collected by the positioning device; This is the speed influence coefficient, with a value range of 0.2-0.4, used to correct the impact of personnel movement speed on filter gain; Let k be the speed at which the person moves, expressed in m / s, which is the actual speed at which the person moves at time k. This represents the maximum permissible movement speed for personnel within the factory, expressed in m / s, which is the maximum movement speed stipulated by factory safety management regulations. Finally, step S35 is executed to converge the identified drifting positioning points to the nearest valid grid cell, a grid cell not currently in the occupied grid list, thus generating a high-precision positioning trajectory.

[0180] To determine the optimal convergence distance, a convergence distance calculation algorithm is used, and the calculation formula is as follows: ,in This represents the optimal convergence distance from the drift point to the valid grid. These are the coordinates of the drift point, i.e., the Cartesian coordinates corresponding to the drift positioning point; These are the coordinates of the center point of the valid grid, i.e., the Cartesian coordinates of the geometric center of the nearest valid grid. This is the convergence coefficient, ranging from 0.9 to 1.1, used to fine-tune the convergence distance to fit the mesh boundary; when the calculated... When the drift point has converged to the legal grid range, the drift point convergence operation is completed.

[0181] In the preferred embodiment, step S4 includes:

[0182] S41. Obtain grid attributes, including hazard attributes and equipment distribution information, through the rule engine; calculate attribute importance using a grid attribute importance evaluation algorithm. The formula is:

[0183] ;

[0184] in For the weight of dangerous attributes, The danger level is... For the weights of device distribution attributes, The importance level of the equipment;

[0185] S42. Determine whether the grid properties meet the preset conditions based on the sensor data;

[0186] S43. If the preset conditions are met, a safety warning or work order dispatch will be triggered;

[0187] S44. Generate dynamic path planning based on grid encoding and grid properties;

[0188] S45. Push route instructions to personnel's mobile terminals; evaluate the push effect using an instruction push success rate calculation algorithm. The formula is:

[0189] ;

[0190] in To indicate the number of successful push notifications. This represents the total number of push notifications. The response time impact coefficient, For terminal response time, For the maximum allowable response time, when At that time, the push notification effect was judged to be good.

[0191] In the preferred scheme, the dynamic path planning generated based on grid codes and grid attributes includes:

[0192] S411. Calculate the grid level distance between the starting point and the ending point using integer bitwise operations;

[0193] S412. The shortest path is generated using the A search algorithm; the heuristic function of the A algorithm is introduced to optimize the algorithm, and the heuristic function is:

[0194] ;

[0195] in, It is the total cost of node n. It is the actual cost from the starting point to node n. It is a heuristic cost estimate from node n to the target node;

[0196] The heuristic cost estimate is calculated using a chosen heuristic function, specifically the Manhattan distance, as shown in the formula:

[0197] ;

[0198] in, The x-axis coordinate of the center point of the current grid node n; The x-axis coordinate of the center point of the grid corresponding to the target grid node; The y-axis coordinate of the center point of the current grid node n; The y-axis coordinate of the center point of the target mesh node;

[0199] The optimized heuristic function is:

[0200] ;

[0201] in For heuristic weighting coefficients, The original Manhattan distance heuristic function, This represents the impact coefficient of grid cost. The cost value of the grid where node n is located. This represents the maximum grid cost value.

[0202] S413. Dynamically adjust path costs based on grid attributes, setting high costs for grids in restricted areas and decreasing costs for grids in risk areas based on distance from the center; calculate the cost of risk area grids using a risk area cost refinement algorithm. The formula is:

[0203] ;

[0204] in The greatest cost to the risk zone The distance from the risk zone grid to the risk center. This serves as a baseline distance for risk impact.

[0205] S414. Set negative costs for critical equipment meshes and prioritize planning through critical equipment meshes;

[0206] S415. Manage grid nodes using open and closed lists to determine the optimal path; calculate management efficiency using a node management efficiency evaluation algorithm. The formula is ,in This represents the number of nodes that have been processed. For the number of nodes in the open list, The number of nodes in the closed list. Based on the processing time, This refers to the actual processing time, when At that time, the node management efficiency is determined to be up to standard.

[0207] Specific implementation process: The implementation process of step S4 is as follows: S41 to S45: S41 obtains the grid attributes of each grid cell through the rule engine. The attributes include hazard attributes and equipment distribution information; in order to clarify the priority of different attributes for business decisions, the grid attribute importance evaluation algorithm is used to calculate the attribute importance. The formula is ,in This represents the weight of the hazard attribute, used to measure the proportion of the hazard attribute in business judgment; Represents the level of hazard attributes, reflecting the degree of danger within the grid; Represents the weight of equipment distribution attributes, used to measure the proportion of influence of equipment distribution attributes; This represents the importance level of the equipment, reflecting the degree of importance of the equipment within the grid.

[0208] Based on real-time data collected by sensors within the factory, S42 determines whether the attributes of the corresponding grid meet the preset business triggering conditions.

[0209] If the grid properties meet the preset conditions, S43 will automatically trigger automated operations, specifically safety alerts or work order dispatch.

[0210] 44. Based on the spatial location and grid attributes corresponding to the grid code, a dynamic path plan is generated. S45. The generated path instructions are pushed to the personnel's mobile terminal; to evaluate the push effect, an instruction push success rate calculation algorithm is used to calculate the push effect. The formula is ,in This represents the number of successful push notifications, and the number of times the terminal successfully received and responded to the notification. Represents the total number of push notifications; This represents the response time impact coefficient, used to correct the impact of terminal response time on push notification performance. Represents the terminal response time, the time interval from receiving the instruction to receiving a response; Represents the maximum allowed response time, the maximum normal response time set by the system for the terminal; when At that time, the command push effect was good.

[0211] "Generate dynamic path planning based on grid encoding and grid attributes" includes detailed steps S411 to S415: S411 calculates the hierarchical distance between the grids corresponding to the starting point and the ending point of the path through integer bit operations of the grid encoding, and quickly derives the spatial distance relationship between the two based on the integer characteristics of the grid encoding.

[0212] S412 uses the A search algorithm to generate the shortest path; to adapt to the factory grid scenario, the A search algorithm heuristic function optimization algorithm is introduced. First, the basic heuristic function formula is defined as follows: ,in (This represents the total cost of node n, used to determine node priority) This represents the actual cost from the starting point to node n, the cost of the path already traveled; This represents the heuristically estimated cost from node n to the target node, predicting the cost of the remaining path. The heuristically estimated cost is calculated using the Manhattan distance heuristic function, as shown in the formula: ,in This represents the x-axis coordinate of the center point of the current grid node n; The x-axis coordinate of the center point of the grid corresponding to the target grid node; This represents the y-axis coordinate of the center point of the current grid node n; This represents the y-axis coordinate of the center point of the target mesh node.

[0213] The optimized heuristic function formula is as follows: ,in Represents the heuristic weight coefficients, used to adjust the influence of the original Manhattan distance heuristic function; Represents the original Manhattan distance heuristic function; This represents the grid cost impact coefficient, used to incorporate the impact of grid cost on the path; The cost value (basic path cost) of the grid where node n is located; This represents the maximum grid cost value (the maximum cost of all grids within the factory). S413 dynamically adjusts path costs based on grid attributes, setting high costs for grids in restricted areas to prevent paths from passing through them; the cost of grids in risk areas decreases with distance from the risk center, and a risk area cost refinement algorithm is used to calculate the cost of grids in risk areas. The formula is ,in This represents the maximum cost of the risk zone, and the cost of the grid where the risk center is located. This represents the distance from the risk zone grid to the risk center; This represents the baseline distance for risk impact, which is the baseline length of the risk impact range set by the system.

[0214] S414 assigns a negative cost to critical equipment meshes, causing the A search algorithm to prioritize paths through these meshes. S415 manages mesh nodes using open and closed lists, dynamically updating the lists and filtering nodes to determine the optimal path; a node management efficiency evaluation algorithm is used to calculate management efficiency. The formula is ,in Represents the number of nodes that have been processed (the total number of nodes that have been transferred from the open list to the closed list). Represents the number of nodes in the open list; Represents the number of nodes in the closed list; Represents the baseline processing time (the standard time for processing the same number of nodes). Represents the actual processing time (the actual processing time consumed); when At that time, the node management efficiency is determined to be up to standard.

[0215] In the preferred scheme, after generating a unique grid code in S1, a grid code fault tolerance verification algorithm is introduced to verify the integrity and correctness of the grid code, including:

[0216] S15. Extract the hierarchy identifier from the mesh code. Longitude offset Latitude offset ;

[0217] S16. Calculate the code check value The formula is ,in The total number of bits in the code. To verify the modulus value;

[0218] S17. The calculated result The code is compared with the built-in check bit. If they match, the code is considered correct. If they do not match, the grid code is regenerated and steps S15-S17 are repeated until the code verification passes, ensuring that the grid code is transmitted without errors in subsequent data association and business processing.

[0219] After generating a unique grid code in step S1, the specific implementation steps of the grid code fault tolerance verification algorithm to verify the integrity and correctness of the grid code are introduced. The overall process revolves around code information extraction, check value calculation, and verification comparison to ensure that the grid code is error-free in subsequent data association and business processing. The specific explanation is as follows: First, step S15 is executed to extract three core components from the generated grid code, namely the hierarchy identifier. Longitude offset Latitude offset These three pieces of information are key to achieving spatial positioning and hierarchical differentiation in grid coding, and are also the basic data for subsequent verification calculations.

[0220] Next, step S16 is executed to calculate the code check value using the formula. The calculation formula is: ,in It serves as a hierarchy identifier in grid coding, used to distinguish the subdivision level to which the grid belongs; This is the longitude offset in grid coding, reflecting the distance the grid is offset from the reference point in the longitude direction; This is the latitudinal offset in grid encoding, reflecting the distance the grid is offset from the reference point in the latitudinal direction; and The coefficient is used to convert the hierarchical identifier, longitude offset, and latitude offset into values ​​of different orders of magnitude, so as to avoid numerical overlap and interference during the addition of the three. This represents the total number of bits in the encoding, and its value is determined by the grid level. The higher the grid level, the more bits are encoded. The larger the value, the range is 32-64; The check value is used to perform a modulo operation on the sum calculated above, generating a check value with a controllable range. .

[0221] Finally, execute step S17, and use the result calculated in S16. The grid code is compared with the check digits provided by the grid code. If they match, the grid code is considered complete and correct. If they do not match, the code is incorrect and needs to be regenerated. Steps S15-S17 are repeated for verification until the code verification passes. This ensures that subsequent operations related to the grid code, such as data association, location calculation, and business logic mapping, will not be affected by the code error.

[0222] In the preferred scheme, after generating a high-precision positioning trajectory in S3, a trajectory similarity analysis algorithm is introduced to verify the consistency of the positioning trajectory within a continuous time period, including:

[0223] S36. Select two adjacent time periods and Within the positioning trajectory, extract the grid-coded sequence on the trajectory respectively. and ,in The number of location points within each time period;

[0224] S37. Calculate the trajectory similarity Sim using the following formula:

[0225] ;

[0226] in The number of identical grid codes in the two sequences. for Length, for The length; and simultaneously introducing the trajectory fluctuation coefficient. The formula is:

[0227] ;

[0228] in for The Middle The center point coordinates of each grid code for The Middle The center point coordinates of each grid code This represents the side length of the current level grid.

[0229] S38. When and If the trajectory is consistent and the positioning data is stable, then step S3 is executed again to correct the positioning and further improve the reliability of the positioning trajectory.

[0230] Specific implementation steps: First, execute step S36, selecting two adjacent time periods from the location records, and recording them as follows: and The two time periods have the same length, ranging from 5 to 10 minutes; extract The grid-coded sequence corresponding to the positioning trajectory within a time period is denoted as Simultaneously extract The grid-coded sequence corresponding to the positioning trajectory within a time period is denoted as ,in This represents the number of location points within each time period. The number of location points in both sequences must be consistent to ensure comparability in subsequent calculations. Next, step S37 is executed to calculate the similarity Sim between the two trajectories using the formula: Sim is the trajectory similarity, which measures the degree of similarity between the location trajectories in two time periods. for and The number of identical grid codes in two grid-coded sequences; for The length, i.e. Number of location points within a time period ; for The length, i.e. Number of location points within a time period This formula quantifies the degree of overlap of trajectories by the ratio of the same number of codes to the total number of codes.

[0231] Simultaneously, a trajectory fluctuation coefficient is introduced. The calculation formula is: ,in The trajectory fluctuation coefficient is used to measure the fluctuation range of the positioning trajectory between two time periods. for The Middle The coordinates of the center point of each grid code The center point's x-axis coordinates are... The y-coordinate of the center point; for The Middle The coordinates of the center point of each grid code The center point's x-axis coordinates are... The y-coordinate of the center point; for and The Middle The Euclidean distance between the corresponding grid center points; This is the sum of the Euclidean distances between the center points of all corresponding grid points in the two sequences; The number of location points within each time period; The edge length of the current level grid is used to standardize the total distance to a fluctuation coefficient related to the grid edge length. Finally, step S38 is executed to set the criteria for trajectory consistency: when the calculated... and If the location trajectories in the two time periods are consistent, the current location data is stable and can be used for subsequent business processing; otherwise... or If the trajectory consistency is not met, the positioning data is unstable and the S3 step needs to be repeated for positioning correction. The positioning data is corrected by means of grid neighborhood topology relationship and personnel movement speed constraint until the trajectory similarity analysis algorithm is executed again and the judgment condition is met, thereby further improving the reliability of the positioning trajectory.

[0232] In the preferred embodiment, after the automated operation is triggered by S4, an operation effect feedback evaluation algorithm is introduced to quantitatively evaluate the execution effect of the automated operation, including:

[0233] S46. Determine the type of automated operation and extract relevant parameters; for safety warning operations, extract the warning response time. Early warning accuracy Work order dispatch operation extracts work order completion rate. Work order completion time ; Path planning operation extracts path deviation rate Path time ;

[0234] S47. Calculate the overall evaluation value of the operation. The formula is:

[0235] ;

[0236] in To determine the number of parameters related to the operation, For the first The weights of each parameter,

[0237] in, Determined based on the type of operation, including safety warning operations. , ;

[0238] Work order dispatching in progress , ;

[0239] In the path planning operation , ;

[0240] For the first The standardized values ​​of each parameter For time-related parameters, smaller The larger the value, the better for parameters such as accuracy and completion rate. , For the first The maximum value of each parameter. For the first The minimum value of each parameter;

[0241] S48. When When the time is right, the automated operation is deemed to be working well; if Then, analyze the reasons for parameter deviations, adjust the business logic mapping relationship in the rule engine, repeat the S4 steps to optimize automated operations, and ensure that the business processing effect meets the factory management requirements.

[0242] The specific explanation is as follows:

[0243] First, execute step S46 to identify the type of automated operation currently triggered. Automated operation types are categorized into three types: safety alert, work order dispatch, and route planning. Extract the corresponding operation-related parameters based on the different operation types. For safety alert operations, extract two parameters: alert response time. The time interval from when the grid properties meet the warning conditions to when the warning information is issued;

[0244] Early warning accuracy The proportion of correct warnings to the total number of warnings;

[0245] The work order dispatch operation extracts two parameters: work order completion rate and work order completion rate. The percentage of completed work orders out of the total number of work orders dispatched, and the work order completion time. The time interval from work order dispatch to work order completion;

[0246] The path planning operation extracts two parameters: path deviation rate and path deviation rate. The proportion of the actual travel path deviation from the planned path to the total length of the planned path, and the path travel time. The time interval from the starting point to the destination along the planned path.

[0247] Next, proceed to step S47 to calculate the overall operational effect evaluation value using the formula. The calculation formula is: ,in This is a comprehensive performance evaluation value used to quantitatively measure the overall execution effectiveness of automated operations. To manage the number of relevant parameters, safety alerts, work order dispatching, and route planning operations. The values ​​are all 2; For the first The weights of each parameter are determined based on the operation type. In a safety warning operation, the weight corresponding to the warning response time is... Weight corresponding to early warning accuracy In the work order dispatching process, the weight corresponding to the work order completion rate. Weight corresponding to work order completion time In path planning operations, the weight corresponding to the path deviation rate. The weight corresponding to the path time ; For the first The standardized values ​​of each parameter are used to convert parameters with different dimensions into uniformly comparable values. The calculation methods are divided into two categories. For time-related parameters, the early warning response time is... Work order completion time Path time The calculation formula is: ,in For the first The actual values ​​of each parameter The smaller, A higher value indicates better parameter performance; for parameters like accuracy and completion rate, the warning accuracy rate... Work order completion rate , That is, the actual value of the parameter is directly used as the standardized value; For the first The maximum value of a parameter, that is, the maximum value that the parameter may take in historical records or preset standards; For the first The minimum value of a parameter, that is, the minimum value that the parameter may take in historical records or preset standards.

[0248] Finally, step S48 is executed to set the threshold for judging the overall operational effect evaluation value: when the calculated value is... If the current automated operation is deemed to be working well and requires no adjustment, then... Then it is necessary to analyze the reasons for the parameter deviations that lead to the low evaluation value, such as the accuracy of the early warning. An excessively low threshold may indicate an unreasonable judgment threshold for grid hazard attributes, or a problem with the work order completion time. Excessive length may be due to improper work order resource matching logic or path deviation rate. An excessively high cost may indicate that the grid cost adjustment rules are not realistic. Adjust the mapping relationship between business logic and grid coding in the rule engine based on the cause of the deviation. Then, re-execute step S4 to trigger automated operations, and re-evaluate the effect using the operation effect feedback evaluation algorithm until... This ensures that the results of business processing meet the needs of factory management.

[0249] Example 2

[0250] Further explanation in conjunction with Example 1, such as Figure 1-3 As shown in the diagram, this system uses the Vue 3 framework for front-end development and Cesium for rendering the Earth, including map display and model rendering, at the underlying level. The back-end uses traditional Java + Spring Boot + Maven. The database is the open-source MySQL.

[0251] This system uses integrated BeiDou grid technology to link and bind models with business data, creating a more realistic twin world.

[0252] After the digital twin factory platform is connected to the Beidou grid code, the accuracy of personnel positioning is improved without adding positioning hardware, reducing project investment costs. At the same time, the grid carries business data, enabling the twin factory to process business and no longer just be a display.

[0253] First, the BeiDou grid is based on the national standard GB / T 39409-2020, the BeiDou Grid Location Code. It divides the Earth into countless grids according to hierarchical levels, and then assigns a code to each grid according to defined rules. This means that any location in the world can be represented by a specific code, with the highest grid level capable of representing 1.5 times the Earth's surface area. 1.5cm grid.

[0254] Once the system integrates a grid, all locations within the factory are contained within that grid. Therefore, the location of each component no longer needs to be represented by latitude and longitude, but rather by different grid codes. For example, a gas storage tank within the factory might be located in the 7th-level BeiDou grid N369243, where this code represents 7.73m. It has a 7.73m grid, but it's also in the level 6 grid N3643, where this code represents 61.84m. A 61.84m grid. See below. Figure 1 As shown.

[0255] The advantage of integrating the system with a grid is that it allows business data to be loaded into the grid, for example, an eight-level grid (0.97). If there is an explosive object within a 0.97m area, we will mark this grid in red to represent a high-risk area. Its blast radius may be the size of a level 6 grid. The level 6 grid in which it is located also has a hazard attribute. When staff enter the grid area, a safety warning can be issued.

[0256] For example, when an exhaust valve is venting gas, once we know the location of the exhaust valve, we naturally know the grid code it belongs to. When gas is being vented, the gas sensing device at the terminal knows which grid the gas is venting into and assigns these grids the property of toxic gas. At this time, no activities are allowed in these areas. After the gas is vented, the toxic gas property is removed from these grids, and production activities can be carried out in these areas again.

[0257] For example, in an emergency dispatch scenario, if personnel need to travel from point A to point B, a traditional digital twin system can only calculate the distance using the location's latitude and longitude. However, latitude and longitude are typically represented as long floating-point numbers, such as (112.4234325435, 34.242342441). Long floating-point operations are slow in computers, and temporary obstacles often appear along the path, making calculations impossible in traditional digital twin systems. In a system integrating BeiDou grid codes, these problems do not exist. Firstly, the generation rules of the BeiDou grid codes themselves contain the meaning of distance. Bitwise operations on integers can determine the number of n-level grids that differentiate between two grids. Secondly, each grid level has a fixed length, making calculations convenient. Furthermore, when obstacles are detected by terminal sensing devices, they are displayed in the system grid. Graying out a grid indicates that it is impassable, allowing for the calculation of the path distance. Figure 2 As shown.

[0258] When planning the route, we use the A-search algorithm, a heuristic search algorithm used to find the shortest path from the starting point to the destination in a graph. The core idea is to use a finite queue to expand the nodes and then select the path with the minimum cost. The formula is as follows:

[0259] ;

[0260] in, It is the total cost of node n. It is the actual cost from the starting point to node n. It is a heuristic cost estimate from node n to the target node.

[0261] The heuristic cost estimate is calculated using a chosen heuristic function; we select the Manhattan distance as the heuristic function. The formula is:

[0262] ;

[0263] in, The x-axis coordinate of the center point of the current grid node n; The x-axis coordinate of the center point of the grid corresponding to the target grid node; The y-axis coordinate of the center point of the current grid node n; The y-axis coordinate of the center point of the grid corresponding to the target grid node.

[0264] 1. Initialization:

[0265] Add the starting point to the open list initially. , Heuristic cost estimation from start to finish, .

[0266] Open set: Contains nodes to be evaluated. Initially, it only contains the starting grid.

[0267] Closed set: Contains nodes that have been evaluated. Initially empty.

[0268] 2. Loop:

[0269] Take the node n with the smallest f(n) from the open list.

[0270] If n is the target node, then construct the path and end.

[0271] Move n from the open list to the closed list.

[0272] For each neighboring node m of n:

[0273] If m is already in the closed list, skip it.

[0274] calculate .

[0275] If m is not in the open list, or the value of g(m) for the new path is smaller:

[0276] Set the parent node of m to n.

[0277] Update g(m) and .

[0278] If m is not in the open list, add m to the open list.

[0279] 3. Path reconstruction:

[0280] Starting from the target node, backtrack along the parent node of each node to the starting point to obtain the final path.

[0281] There is a key parameter in the algorithm , representing the cost of moving from grid n to grid m. Since m and n are adjacent, it is actually the cost of moving forward one grid. Normally, we set the movement cost between grids to 1: this is when the grid does not belong to the restricted area (gray grid in the image above) nor the risk area (red grid in the image above). First, we set the movement cost of the restricted area grid to 1000, which is a very large value. Substituting this into the algorithm will prevent such grids from being included in the path. Then, using the Manhattan distance mentioned above, we obtain the distance from each risk area grid M to the grid at the exact center of the risk point. Then use The resulting values ​​are used as the cost of each risk zone grid. This ensures that the cost of risk zone grids farther from the center is smaller, which aligns with real-world logic. This, to some extent, corrects the accuracy of route planning through the algorithm. Then, reward values ​​should be set at certain key points within the factory area to encourage paths to pass through these points. For example, the cost of a mesh containing a fire extinguisher is -100, the cost of a fire blanket is -200, and the cost of a smoke mask is 400. The reason for this design is that in an escape scenario, objects that can protect oneself take priority over fire extinguishing equipment.

[0282] Meanwhile, the grid can carry not only spatial data but also temporal data, and is called the BeiDou spatiotemporal grid. For example, if a flammable material was placed in a grid three months ago and the grid was marked with a flammable attribute, and a fire source operation needs to be started five grids of level 9 away from this location, it is necessary to check whether the grids around the fire source point have a flammable attribute. If it is found that a flammable attribute was added to this grid three months ago and has not been deleted, it is determined that the flammable material has been there and has not been moved. Then, the fire source operation is terminated and the staff is prompted to go to the corresponding grid location to check for flammable materials.

[0283] Example 3

[0284] like Figure 3 As shown, the implementation process is as follows: In the system, all grids are constructed using the centerline element of the Cesium base. The grid is a cube with a side length of 0.97m. The coordinates of the first time the personnel positioning beacon passes through are used as the reference point. If the first set of coordinates is offset into the wall, the reference point is directly corrected and moved to the adjacent normal grid. Then, each subsequent set of coordinates is compared with the previous one to determine whether the number of grids that have changed is greater than 2. If it is greater, they converge to a single grid in the same direction.

[0285] The BeiDou grid code can assist in correction because, under an 8-level grid, each person transmits positioning data back once per second via their positioning card. Considering that an individual's movement distance in one second will not exceed one meter, when a person drifts from one second to the next, the distance exceeds one grid, the system will automatically shrink back to one grid. At the same time, people are not allowed to be positioned in grids occupied by buildings or obstacles. This can avoid most positioning drift scenarios and achieve assisted positioning.

[0286] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A business processing method of a smart factory platform based on a Beidou grid code, characterized by: The method includes: S1. The factory's physical space is divided into multi-level grids using BeiDou grid codes to generate unique grid codes; the grid codes use integer format to represent location information. S2. Bind equipment sensor data, personnel positioning data, and business logic data with grid coding to construct a spatiotemporal joint index table; S3. Real-time correction of positioning data is performed using grid neighborhood topology and personnel movement speed constraints to generate high-precision positioning trajectories; S4. Map business logic to grid code through the rules engine, and trigger automated operations based on grid attributes; Step S1 involves multi-level gridding of the factory's physical space using BeiDou grid codes, including: S11. Based on the functions of the factory area, an octree partitioning algorithm is used to generate multi-level grid cells; each grid cell corresponds to a different side length; a grid partitioning accuracy optimization algorithm is introduced to calculate the matching degree between the grid cell side length and the accuracy required by the regional business needs. The formula is: ； in For precision weighting system, To achieve the minimum positioning accuracy required for regional business needs. The side length of the grid cell. To ensure that a single grid cell effectively covers the business area, For the total area of ​​the region, when At that time, it is determined that the mesh subdivision accuracy meets the business requirements; S12. Assign a unique code to each grid cell; the code includes the level identifier, longitude offset, and latitude offset; S13. Dynamically adjust the grid level according to the regional function. High-resolution grid is used in passage areas, medium-resolution grid is used in storage areas, and low-resolution grid is used in open areas. S14. Generate an occupied grid list, recording the grid cells occupied by buildings and obstacles; use a grid occupancy calculation algorithm to determine the probability that a grid is occupied. The formula is: ； in The number of times the sensor detects obstacles within the grid. For sensor detection weights, Mark the number of times obstacles exist within the grid on the factory map. Label the map with weights. For the total number of detections and annotations, when When this happens, add the grid to the list of occupied grids; Step S4 includes: S41. Obtain grid attributes, including hazard attributes and equipment distribution information, through the rule engine; calculate attribute importance using a grid attribute importance evaluation algorithm. The formula is: ； in For the weight of dangerous attributes, The danger level is... For the weights of device distribution attributes, The importance level of the equipment; S42. Determine whether the grid properties meet the preset conditions based on the sensor data; S43. If the preset conditions are met, a safety warning or work order dispatch will be triggered; S44. Generate dynamic path planning based on grid encoding and grid properties; S45. Push route instructions to personnel's mobile terminals; evaluate the push effect using an instruction push success rate calculation algorithm. The formula is: ； in To indicate the number of successful push notifications. This represents the total number of push notifications. The response time impact coefficient, For terminal response time, For the maximum allowable response time, when At that time, the push notification effect was judged to be good.

2. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: Step S2 involves binding device sensor data, personnel positioning data, and business logic data with grid coding, including: S21. Acquire equipment sensor data, including environmental parameters and equipment status; calculate data reliability using a sensor data reliability assessment algorithm. The formula is: ； in For a single sensor data acquisition, The average value of n collected data points. For the number of data collections, The maximum allowable deviation value for the sensor, when When the data is deemed reliable, it is used for subsequent binding operations; S22. Obtain personnel positioning data, including the coordinate information of continuous positioning points; S23. Associate business logic data with grid coding to form a rule set; S24. Construct a spatiotemporal joint index table to store the mapping relationship between grid coding and sensor data, positioning data and business logic data; S25. Enables rapid retrieval and dynamic updating of multi-source data through index tables; The retrieval response time of the index table is calculated using a retrieval efficiency optimization algorithm. The formula is: ； in For the basic retrieval time of the index table, For data correlation coefficient, This refers to the amount of data retrieved in a single search. For data transmission bandwidth, This formula optimizes the index table structure to improve the index table query hit rate. .

3. The business processing method of a smart factory platform based on Beidou grid codes according to claim 2, characterized in that: Step S24 includes: S241. For each grid code, associate the corresponding sensor data and positioning data; S242. Generate a rule set based on business logic data, including security rules and scheduling rules; calculate the rule validity using a rule set validity evaluation algorithm. The formula is: ； in For the number of times the rule is triggered, To trigger frequency weights, The number of times the rule is correctly executed. To implement correctness weights, For the total number of rules, when When the rule set is valid, it is determined that the rule set is valid. S243. The mapping relationship between grid codes and rule sets is stored in a database; S244. When sensor data is updated, the mapping relationship in the index table is dynamically adjusted; the sensitivity adjustment algorithm is used to calculate the adjusted sensitivity. The formula is: ； in Adjust the number of times for the mapping relationship. For the number of times sensor data is updated, To adjust the coefficient, when At that time, the determination of the mapping relationship adjustment is sensitive; S245. Support cross-level business association queries through index tables.

4. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: Step S3 includes: S31. Obtain the grid code of continuous personnel positioning points and calculate the number of grid spans between adjacent positioning points; S32. If the number of crossings exceeds the preset speed constraint threshold, it is determined to be positioning drift; a dynamic speed threshold calculation algorithm is introduced, the formula is: ； in As the speed constraint threshold, This represents the side length of the current level grid. This is the slope influence coefficient. The slope value of the area where the location point is located. This represents the maximum slope value within the factory area. This refers to the time interval for location data collection. S33. Filter out illegal location points by combining the occupied grid list; S34. Smooth the positioning data using Kalman filtering; introduce a filter gain optimization algorithm to calculate the Kalman filter gain. The formula is: ； in Let the prior error covariance be at time k. For the observation matrix, To observe the noise covariance, For speed influence coefficient, Let k be the speed at which people move. This refers to the maximum permissible movement speed of personnel within the factory. S35. Converge the drifting positioning points to the nearest valid grid to generate a high-precision positioning trajectory; use a convergence distance calculation algorithm to determine the optimal convergence distance from the drifting points to the valid grid. The formula is: ； in Here are the coordinates of the drift point. The coordinates of the center point of the valid grid. Let be the convergence coefficient, when At that time, the drift point convergence is complete.

5. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: The dynamic path planning process based on grid encoding and grid properties includes: S411. Calculate the grid level distance between the starting point and the ending point using integer bitwise operations; S412. The shortest path is generated using the A search algorithm; the heuristic function of the A algorithm is introduced to optimize the algorithm, and the heuristic function is: ； in, It is the total cost of node n. It is the actual cost from the starting point to node n. It is a heuristic cost estimate from node n to the target node; The heuristic cost estimate is calculated using a chosen heuristic function, specifically the Manhattan distance, as shown in the formula: ； in, The x-axis coordinate of the center point of the current grid node n; The x-axis coordinate of the center point of the grid corresponding to the target grid node; The y-axis coordinate of the center point of the current grid node n; The y-axis coordinate of the center point of the target mesh node; The optimized heuristic function is: ； in For heuristic weighting coefficients, The original Manhattan distance heuristic function, This represents the impact coefficient of grid cost. The cost value of the grid where node n is located. This represents the maximum grid cost value. S413. Dynamically adjust path costs based on grid attributes, setting high costs for grids in restricted areas and decreasing costs for grids in risk areas based on distance from the center; calculate risk area grid costs using a risk area cost refinement algorithm. The formula is: ； in The greatest cost to the risk zone The distance from the risk zone grid to the risk center. This serves as a baseline distance for risk impact. S414. Set negative costs for critical equipment meshes and prioritize planning through critical equipment meshes; S415. Manage grid nodes using open and closed lists to determine the optimal path; calculate management efficiency using a node management efficiency evaluation algorithm. The formula is ,in This represents the number of nodes that have been processed. For the number of nodes in the open list, The number of nodes in the closed list. Based on the processing time, This refers to the actual processing time, when At that time, the node management efficiency is determined to be up to standard.

6. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: in After S1 generates a unique grid code, a grid code fault tolerance verification algorithm is introduced to verify the integrity and correctness of the grid code, including: S15. Extract the hierarchy identifier from the mesh code. Longitude offset Latitude offset ; S16. Calculate the code check value The formula is ,in The total number of bits in the code. To verify the modulus value; S17. The calculated result The code is compared with the built-in check bit. If they match, the code is considered correct. If they do not match, the grid code is regenerated and steps S15-S17 are repeated until the code verification passes, ensuring that the grid code is transmitted without errors in subsequent data association and business processing.

7. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: After generating a high-precision positioning trajectory using S3, a trajectory similarity analysis algorithm is introduced to verify the consistency of the positioning trajectories within a continuous time period, including: S36. Select two adjacent time periods and Within the positioning trajectory, extract the grid-coded sequence on the trajectory respectively. and ,in The number of location points within each time period; S37. Calculate the trajectory similarity Sim using the following formula: ； in The number of identical grid codes in the two sequences. for Length, for The length; and simultaneously introduce the trajectory fluctuation coefficient. The formula is: ； in for The Middle The center point coordinates of each grid code for The Middle The center point coordinates of each grid code This represents the side length of the current level grid. S38. When and If the trajectory is consistent and the positioning data is stable, then step S3 is executed again to correct the positioning and further improve the reliability of the positioning trajectory.

8. The business processing method of a smart factory platform based on Beidou grid codes according to claim 1, characterized in that: in After S4 triggers an automated operation, an operation effect feedback evaluation algorithm is introduced to quantitatively evaluate the execution effect of the automated operation, including: S46. Determine the type of automated operation and extract relevant parameters; for safety warning operations, extract the warning response time. Early warning accuracy Work order dispatch operation extracts work order completion rate. Work order completion time Path planning operation extracts path deviation rate Path time ; S47. Calculate the overall evaluation value of the operation. The formula is: ； in To determine the number of parameters related to the operation, For the first The weights of each parameter, in, Determined based on the type of operation, including safety warning operations. , ; Work order dispatching in progress , ; In the path planning operation , ; For the first The standardized values ​​of each parameter For time-related parameters, smaller The larger the value, the better for parameters such as accuracy and completion rate. , For the first The maximum value of each parameter. For the first The minimum value of each parameter; S48. When When the time is right, the automated operation is deemed to be working well; if Then, analyze the reasons for parameter deviations, adjust the business logic mapping relationship in the rule engine, repeat the S4 steps to optimize automated operations, and ensure that the business processing effect meets the factory management requirements.

Images