A data processing method based on industrial Internet data platform

Abstract

The present invention relates to the field of industrial Internet technology, and in particular to a data processing method based on an industrial Internet data platform. The method comprises the following steps: using the industrial Internet data platform to process industrial production context and generate real-time industrial production context data; constructing a production process constraint knowledge graph based on the real-time industrial production context data; using the production process constraint knowledge graph to match order priorities and generate order priority allocation data; performing production plan conflict detection based on the order priority allocation data to obtain subtask production conflict data; performing flexible scheduling processing based on the subtask production conflict data to generate conflict-free production subtask network data; performing key task path chain identification based on the conflict-free production subtask network data to obtain final production plan data. The present invention eliminates conflicts in production plans and significantly improves production efficiency through flexible scheduling of task chains.

Description

Technical Field

[0001] The present invention relates to the field of industrial Internet technology, and in particular to a data processing method based on an industrial Internet data platform. Background Art

[0002] With the rapid development of the industrial Internet, data has become a key resource for corporate competitiveness. As a product of the deep integration of new-generation information technology and manufacturing, the industrial Internet is driving the transformation of the manufacturing industry towards digitalization, networking and intelligence. Through the application of industrial Internet technology, production planning and scheduling are gradually moving towards intelligence and automation. As its core component, the industrial Internet data platform undertakes key tasks such as the collection, storage, processing and analysis of massive industrial data. By integrating various industrial equipment, systems and applications, these platforms provide enterprises with solutions for real-time data collection, transmission and storage, thereby improving production efficiency and decision-making accuracy, and promoting the overall upgrade of the manufacturing industry. However, the traditional data processing method based on the industrial Internet data platform for production planning and scheduling mainly relies on manual experience, lacks flexibility and foresight, and is difficult to adapt to complex and changing market demands; in a complex production environment, due to production plan conflicts, it is impossible to quickly generate executable conflict-free plans, resulting in low production efficiency. Summary of the invention

[0003] Based on this, the present invention provides a data processing method based on an industrial Internet data platform to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a data processing method based on an industrial Internet data platform includes the following steps:

[0005] Step S1: Use the industrial Internet data platform to collect real-time data on the production site to obtain the original industrial data stream; perform industrial production context processing based on the original industrial data stream to generate real-time industrial production context data;

[0006] Step S2: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data; perform production capacity analysis on the real-time industrial production resource data to obtain a production capacity dynamic index; construct a production capacity matrix for the real-time industrial production context data using the production capacity dynamic index to obtain initial production capacity matrix data;

[0007] Step S3: construct a production process constraint knowledge graph based on real-time industrial production context data; use the production process constraint knowledge graph to perform production constraint correction processing on the initial production capacity matrix data to generate optimized production capacity matrix data; perform comprehensive production capacity analysis based on the optimized production capacity matrix data to generate comprehensive production capacity data;

[0008] Step S4: performing demand production forecasting based on the comprehensive production capacity data to obtain demand production forecasting data; performing order priority matching based on the demand production forecasting data to generate order priority allocation data;

[0009] Step S5: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data to generate preliminary subtask unit data; construct production subtask network data based on the preliminary subtask unit data; perform production plan conflict detection on the production subtask network data to obtain subtask production conflict data; use the subtask production conflict data to perform flexible scheduling processing on the production subtask network data to generate conflict-free production subtask network data;

[0010] Step S6: Identify the critical task path chain based on the conflict-free production subtask network data, and perform insertion planning of non-critical path tasks to generate task scheduling plan data; perform low-risk production plan execution processing on the task scheduling plan data to obtain the final production plan plan data.

[0011] The present invention collects production site data in real time through the industrial Internet data platform, ensuring the timeliness and accuracy of the data. The industrial production context processing not only integrates these raw data, but also reveals the internal connections and flow patterns in the production process. The generated real-time industrial production context data helps to deeply understand the production status and process, and sort out the key links and logical relationships of the production process. The real-time industrial production context data is analyzed for resource nodes, and the generated real-time industrial production resource data helps to quickly locate resource bottlenecks, and can accurately identify and quantify the resource input of each production link. The dynamic index is obtained through production capacity analysis, which can reflect the changes in production capacity in real time. The production capacity matrix is ​​constructed to comprehensively characterize the capacity status of the production system. The production process constraint knowledge map is constructed to systematically express various constraint relationships in the production process. The ability to understand and control complex production environments is enhanced. The production capacity matrix is ​​corrected using the knowledge map to ensure the feasibility and rationality of the production plan, and reflects the optimal production capacity layout under all actual constraints. Demand forecasting based on comprehensive production capacity can more accurately estimate future production demand and provide forward-looking guidance for resource allocation and production scheduling. Order priority matching arranges the production sequence reasonably based on the prediction results to ensure the timely completion of high-priority orders, thereby improving customer satisfaction and market response speed. Task decoupling processing can split complex orders into executable subtask units, refine and optimize the production task structure, and generate preliminary subtask unit data for flexible management and scheduling. The construction of the production subtask network and the conflict detection mechanism effectively avoid time and resource conflicts in the production plan, ensuring the smooth and efficient production process. Critical task path identification helps to grasp the main line of the production plan, which not only optimizes the task execution sequence, but also maximizes the utilization of production resources and improves production efficiency. Low-risk production plan execution processing ensures the stable implementation of the production plan. The final production plan data not only comprehensively considers efficiency, cost and risk, but also has a high degree of executableness, in order to achieve the goal of intelligent and refined production management. Therefore, a data processing method based on an industrial Internet data platform of the present invention utilizes the industrial Internet data platform to collect production site data in real time and integrate multi-source heterogeneous data; through production resource node analysis and equipment and personnel capability assessment, a production process constraint knowledge graph is constructed, and this graph is used to perform constraint correction and quantitative analysis on production capacity. Combining historical data with real-time data, a prediction model is used to perform demand forecasting, thereby realizing dynamic adjustment and optimization of production plans, and adopting strategies such as time constraint relaxation, resource substitution, and constraint parameter adjustment to eliminate conflicts and re-optimize production plans, thereby ensuring the conflict-free and efficient execution of production plans.

[0012] Preferably, step S1 comprises the following steps:

[0013] Step S11: Use the industrial Internet data platform to collect real-time data from the production site to obtain the original industrial data stream;

[0014] Step S12: performing collaborative fusion of heterogeneous data according to the original industrial data stream to generate fused original industrial data;

[0015] Step S13: performing adaptive data cleaning and filtering on the fused original industrial data to obtain cleaned industrial production data;

[0016] Step S14: performing timestamp calibration processing based on the original industrial production data to generate industrial production calibration data;

[0017] Step S15: Perform data association aggregation processing according to the industrial production calibration data, and perform industrial production context processing to generate real-time industrial production context data.

[0018] The present invention uses the industrial Internet data platform to collect real-time data on the production site, and can obtain the original industrial data stream comprehensively, accurately and timely. The original industrial data stream is subjected to the collaborative fusion of heterogeneous data, which can effectively integrate multi-source heterogeneous data from different devices and systems. This fusion process eliminates the differences in data format and structure, and realizes the unified expression and storage of data. Adaptive data cleaning and filtering can intelligently identify and process noise, outliers and redundant information in the original data, significantly improve the data quality, and remove interference factors that affect the accuracy of the analysis results. The adaptive characteristics enable the cleaning process to automatically adjust the processing strategy according to different data characteristics, which not only ensures the effect of data cleaning, but also improves the processing efficiency. The timestamp calibration process of industrial production data solves the problem of time asynchrony of multi-source data. Accurate time synchronization is crucial for analyzing the cause-effect relationship and timing pattern in the production process. The calibrated data can accurately reflect the actual occurrence order and duration of production events. Data association aggregation processing can discover and establish logical relationships between different data, and organize discrete data points into meaningful data sets. Subsequent industrial production context processing further sorts out the key links and process logic of the production process.

[0019] Preferably, step S2 comprises the following steps:

[0020] Step S21: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data, wherein the real-time industrial production resource data includes industrial production equipment node data and staff node data;

[0021] Step S22: dynamically updating equipment parameters of a preset production equipment digital model through industrial production equipment node data to generate a dynamic production equipment digital model;

[0022] Step S23: quantifying the equipment production capacity of the dynamic production equipment digital model to obtain real-time equipment production capacity data;

[0023] Step S24: extracting personnel skills according to the personnel node data, and matching personnel skills through a preset personnel skill model library to generate personnel skill data; performing production skill proficiency assessment processing on the personnel skill data through the personnel node data to generate personnel production capacity data;

[0024] Step S25: performing production capacity processing on the real-time industrial production resource data based on the personnel production capacity data and the real-time equipment production capacity data to obtain a production capacity dynamic index;

[0025] Step S26: label the real-time industrial production context data with production capacity through the production capacity dynamic index, and construct a production capacity matrix to obtain initial production capacity matrix data.

[0026] The present invention can accurately identify and quantify key resource elements in the production process by performing production resource node analysis on real-time industrial production context data, dynamically reflect the real-time status of production resources, and help to timely discover resource bottlenecks and optimize resource allocation. The preset digital model of production equipment is dynamically updated using industrial production equipment node data, ensuring the high consistency between the digital model and the actual equipment status, and improving the accuracy and efficiency of equipment management. Quantitative processing of equipment production capacity on the dynamic production equipment digital model helps to accurately evaluate the contribution of equipment to the overall production capacity and support data-based equipment investment and upgrade decisions. Personnel skills are extracted and matched through staff node data to achieve an accurate assessment of human resource capabilities. The generated staff skill data fully reflects the professional capability structure of employees. Based on the comprehensive processing of production capacity data of personnel and equipment, the obtained production capacity dynamic index fully reflects the capacity status of the overall production system. The real-time industrial production context data is labeled and matrixed by the production capacity dynamic index, achieving accurate mapping of production capacity and production process, and the initial production capacity matrix data intuitively shows the capacity distribution of each link in the production process.

[0027] Preferably, step S25 comprises the following steps:

[0028] Step S251: Evaluate the equipment load through the industrial production equipment node data to generate production equipment load index data;

[0029] Step S252: Perform equipment availability analysis based on production equipment load index data to generate an equipment availability index;

[0030] Step S253: Performing personnel availability analysis through the personnel node data to generate a personnel availability index;

[0031] Step S254: using the equipment availability index and the personnel availability index to perform spatiotemporal production node status evaluation on the real-time industrial production context data to generate spatiotemporal production status evaluation data;

[0032] Step S255: Perform productivity field processing according to the spatiotemporal production status evaluation data to obtain industrial productivity distribution data;

[0033] Step S256: Perform capacity decay analysis on the industrial productivity field data based on the personnel production capacity data and the real-time equipment production capacity data to generate capacity decay data;

[0034] Step S257: performing capacity compensation processing according to the capacity decay data to obtain productivity capacity compensation data;

[0035] Step S258: Calculate the production capacity dynamic index of the industrial productivity distribution data using the productivity capacity compensation data to generate the production capacity dynamic index.

[0036] The present invention uses industrial production equipment node data to evaluate equipment load, which can accurately reflect the current working status of each device, help identify the high and low equipment utilization rate, and discover potential equipment overload or idle problems. Equipment availability analysis is performed based on production equipment load index data, which fully reflects the actual available status of the equipment, takes into account factors such as equipment load and failure rate, and helps to improve the accuracy and executability of production plans. Personnel availability analysis is performed through staff node data, which accurately reflects the real-time status of human resources, takes into account factors such as working hours and skill matching, and helps to improve personnel utilization and work efficiency. The equipment and personnel availability index is used to evaluate the spatiotemporal production node status of real-time industrial production context data, which fully reflects the dynamic operation status of the production system. The productivity field is processed based on the spatiotemporal production status evaluation data, which intuitively shows the distribution of production capacity in the entire production system, and helps to understand the spatial distribution characteristics of production capacity. Based on personnel and equipment production capacity data, the capacity decay analysis of industrial productivity field data reflects the dynamic characteristics of production capacity changing with time and space, and helps to predict the changing trend of production capacity. Capacity compensation processing is performed based on capacity decay data, which helps to balance the fluctuation of production capacity, improve the stability and reliability of the production system, and ensure the continuous achievement of production goals. The production capacity dynamic index is calculated by using the production capacity compensation data to calculate the industrial productivity distribution data, which fully reflects the real-time capacity status of the production system. This dynamic index comprehensively considers multiple factors such as equipment, personnel, time and space distribution, and supports more accurate production management and resource allocation.

[0037] Preferably, step S3 comprises the following steps:

[0038] Step S31: Identify production process elements on real-time industrial production context data to generate production process element data;

[0039] Step S32: extracting constraints based on the production process element data to generate constraint structure data, wherein the constraint structure data includes production process flow, quality control standards and safety specifications; constructing a production process constraint knowledge graph based on the constraint structure data;

[0040] Step S33: Mapping the initial production capacity matrix data to production constraints through the production process constraint knowledge graph to generate production constraint mapping data;

[0041] Step S34: performing constraint condition association analysis on the production constraint mapping data, and performing constraint condition priority sorting to generate constraint priority data;

[0042] Step S35: dynamically adjusting the constraint conditions of the production constraint mapping data through the constraint priority data to obtain actual production capacity constraint data;

[0043] Step S36: using the actual production capacity constraint data to perform actual capacity value correction processing on the initial production capacity matrix data to obtain optimized production capacity matrix data;

[0044] Step S37: Calculate the comprehensive production capacity based on the optimized production capacity matrix data to generate comprehensive production capacity data.

[0045] The present invention identifies production process elements for real-time industrial production context data, comprehensively reflects the key components of the production system, helps to deeply understand the structure and characteristics of the production process, and ensures the comprehensiveness and accuracy of constraint analysis. Constraints are extracted according to production process element data, various limiting factors in the production process are sorted out, and the constructed production process constraint knowledge map intuitively displays these constraints in the form of a map, reflecting the complex relationship between the constraints. The initial production capacity matrix data is mapped through the production process constraint knowledge map, and the abstract constraints are associated with specific production capacity data. The production constraint mapping data is associated and prioritized, and the importance of different constraints is clarified, which helps to focus on key constraints in a complex production environment and improves the efficiency and accuracy of decision-making. The production constraint mapping data is dynamically adjusted through constraint priority data, which reflects the real constraint status of the production system in actual operation, can adapt to changes in the production environment, and ensure that the constraint analysis is always consistent with the actual situation. The initial production capacity matrix data is corrected using the actual production capacity constraint data, which more accurately reflects the actual capacity of the production system, eliminates the gap between theoretical capacity and actual capacity, and reduces the deviation in plan execution. The comprehensive production capacity is calculated based on the optimized production capacity matrix data, which fully reflects the overall capacity level of the production system.

[0046] Preferably, step S4 comprises the following steps:

[0047] Step S41: Acquire historical production order data and real-time production order data; perform feature engineering processing on the historical production order data to generate historical production order feature data;

[0048] Step S42: performing demand production forecasting on the historical production order feature data by using a preset autoregressive difference moving average model to obtain demand production forecasting data;

[0049] Step S43: using the comprehensive production capacity data to dynamically adjust the demand production forecast data to generate revised demand production forecast data;

[0050] Step S44: performing order urgency processing on the real-time production order data to generate order production urgency data;

[0051] Step S45: Calculate the order priority of the order production urgency data by modifying the demand production forecast data to generate order priority allocation data.

[0052] The present invention can extract key features that affect production demand from massive data by integrating historical production order data with real-time production order data and performing feature engineering on historical data. The demand production forecast is performed on the historical production order feature data through a preset autoregressive difference moving average model. This forecasting method based on time series analysis takes into account the trend, seasonality and cyclical characteristics of historical data and can capture the dynamic characteristics of demand changes. The demand production forecast data is dynamically adjusted using comprehensive production capacity data to be more in line with actual production capacity. The revised forecast data can better balance demand and supply and avoid the risk of overproduction or underproduction. The real-time production order data is processed according to the order urgency, which reflects the time sensitivity of different orders and helps to identify key orders that need to be processed first. By clarifying the urgency of the order, the enterprise can respond to customer needs more flexibly. By correcting the demand production forecast data, the order production urgency data is calculated for order priority, which comprehensively considers the forecast demand and order urgency, and can maintain the stability of production while meeting customer needs.

[0053] Preferably, step S5 comprises the following steps:

[0054] Step S51: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data, and perform subtask process module mapping to generate preliminary subtask unit data;

[0055] Step S52: clustering the demand clusters according to the preliminary subtask unit data, and constructing the task network to obtain the production subtask network data;

[0056] Step S53: Prioritize the production subtask network data using the order priority allocation data to obtain priority production subtask network data;

[0057] Step S54: Performing production plan conflict detection on the priority production subtask network data by optimizing the production capacity matrix data and the production process constraint knowledge graph to obtain subtask production conflict data, wherein the subtask production conflict data includes subtask time conflict data, subtask production resource conflict data and subtask production constraint violation data;

[0058] Step S55: Utilize the subtask production conflict data to perform flexible scheduling processing on the priority production subtask network data to generate conflict-free production subtask network data.

[0059] The present invention uses the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data, splits the complex production orders into manageable small units, improves the flexibility and controllability of the production tasks, and facilitates refined management. The preliminary subtask unit data is clustered and the task network is constructed to clearly show the logical relationship and dependency between the subtasks. This network structure helps to identify critical paths and parallel tasks. The production subtask network data is prioritized by the order priority allocation data, reflecting the importance and urgency of different subtasks. This priority labeling mechanism ensures that the key tasks are processed in a timely manner, which helps to reasonably allocate production resources under limited resources. The production plan conflict detection is performed on the priority production subtask network data, which fully reflects the potential production problems. This conflict detection covers multiple dimensions such as time, resources and constraints, and can timely discover and prevent bottlenecks and problems in the production process. The priority production subtask network data is flexibly scheduled, providing an executable optimization solution for actual production, which can flexibly respond to various production conflicts and maximize resource utilization efficiency while ensuring the realization of production goals. The conflict-free subtask network ensures the smooth execution of production plans, reduces the risk of production interruptions and resource waste, and improves overall production efficiency.

[0060] Preferably, step S6 comprises the following steps:

[0061] Step S61: identifying the critical task path chain according to the conflict-free production subtask network data, and generating critical path task chain data;

[0062] Step S62: Performing operation time window processing according to the critical path task chain data to generate critical operation path timing data;

[0063] Step S63: Perform non-critical path task insertion planning on the conflict-free production subtask network data through the critical operation path timing data to generate task scheduling plan data;

[0064] Step S64: Perform a scheduling risk assessment on the task scheduling data, and perform low-risk production plan execution processing to obtain final production plan data.

[0065] The present invention identifies the critical task path chain based on the conflict-free production subtask network data, which helps to highlight the critical task sequences that directly affect the total production cycle. By processing the critical path task chain data through the operation time window, a clear time limit is set for each critical task to ensure that these tasks are completed within the specified time, avoid delays on the critical path, and improve the accuracy and execution of the production plan. Through the critical operation path timing data, non-critical path task insertion planning is performed on the conflict-free production subtask network data to ensure that the time window on the non-critical path is maximized on the premise of ensuring the smooth execution of the critical tasks, increase the flexibility of the production plan, and optimize the allocation and use of resources, thereby improving production efficiency. The task scheduling plan data is evaluated for the scheduling plan risk, ensuring that the production plan has undergone comprehensive risk considerations and optimization before execution. Risk assessment helps to identify potential execution obstacles and risk points in advance, while the low-risk production plan execution processing ensures the stability and success rate of the production plan by avoiding or reducing these risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic diagram of the steps of a data processing method based on an industrial Internet data platform of the present invention;

[0067] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0068] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0069] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0070] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0071] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0072] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0073] To achieve this, please refer to Figures 1 to 3 The present invention provides a data processing method based on an industrial Internet data platform, comprising the following steps:

[0074] Step S1: Use the industrial Internet data platform to collect real-time data on the production site to obtain the original industrial data stream; perform industrial production context processing based on the original industrial data stream to generate real-time industrial production context data;

[0075] Step S2: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data; perform production capacity analysis on the real-time industrial production resource data to obtain a production capacity dynamic index; construct a production capacity matrix for the real-time industrial production context data using the production capacity dynamic index to obtain initial production capacity matrix data;

[0076] Step S3: construct a production process constraint knowledge graph based on real-time industrial production context data; use the production process constraint knowledge graph to perform production constraint correction processing on the initial production capacity matrix data to generate optimized production capacity matrix data; perform comprehensive production capacity analysis based on the optimized production capacity matrix data to generate comprehensive production capacity data;

[0077] Step S4: performing demand production forecasting based on the comprehensive production capacity data to obtain demand production forecasting data; performing order priority matching based on the demand production forecasting data to generate order priority allocation data;

[0078] Step S5: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data to generate preliminary subtask unit data; construct production subtask network data based on the preliminary subtask unit data; perform production plan conflict detection on the production subtask network data to obtain subtask production conflict data; use the subtask production conflict data to perform flexible scheduling processing on the production subtask network data to generate conflict-free production subtask network data;

[0079] Step S6: Identify the critical task path chain based on the conflict-free production subtask network data, and perform insertion planning of non-critical path tasks to generate task scheduling plan data; perform low-risk production plan execution processing on the task scheduling plan data to obtain the final production plan plan data.

[0080] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a data processing method based on an industrial Internet data platform of the present invention. In this embodiment, the data processing method based on an industrial Internet data platform includes the following steps:

[0081] Step S1: Use the industrial Internet data platform to collect real-time data on the production site to obtain the original industrial data stream; perform industrial production context processing based on the original industrial data stream to generate real-time industrial production context data;

[0082] In an embodiment of the present invention, the sensor network deployed by the industrial Internet data platform collects production site data in real time, such as equipment operating status, material consumption, product quality, etc., to form an original industrial data stream. Through pre-processing operations such as data cleaning and filtering, noise data is eliminated, and the original data is converted into real-time industrial production context data containing time, location, operation and other information in combination with the production process and equipment association relationship. For example, the platform collects temperature, pressure, and vibration data of a certain processing center, as well as the product ID and process information being processed by the processing center. Through data preprocessing and association analysis, the real-time processing status, processing parameters, environmental parameters and other information of the product in the processing center can be generated to form the real-time industrial production context data of the product in the process.

[0083] Step S2: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data; perform production capacity analysis on the real-time industrial production resource data to obtain a production capacity dynamic index; construct a production capacity matrix for the real-time industrial production context data using the production capacity dynamic index to obtain initial production capacity matrix data;

[0084] In an embodiment of the present invention, for example, based on the real-time industrial production context data of the product in the processing center, the equipment ID, operator ID, raw material ID and other information used in the link can be extracted, and combined with the equipment information library, personnel information library, material information library, etc., real-time industrial production resource data containing information such as equipment status, personnel skills, and material inventory is generated. Then, combined with historical data, equipment parameters, personnel skills and other information, the production capacity of each production link is simulated and evaluated, for example, using queuing network models, discrete event simulation and other methods to obtain a dynamic index of production capacity. According to the dynamic index of production capacity, the real-time industrial production context data is weighted and processed to construct the initial production capacity matrix data. For example, each element in the matrix represents the maximum number of products that can be produced in a specific process within a specific time period.

[0085] Step S3: construct a production process constraint knowledge graph based on real-time industrial production context data; use the production process constraint knowledge graph to perform production constraint correction processing on the initial production capacity matrix data to generate optimized production capacity matrix data; perform comprehensive production capacity analysis based on the optimized production capacity matrix data to generate comprehensive production capacity data;

[0086] In an embodiment of the present invention, a knowledge graph of production process constraints is constructed based on real-time industrial production context data, combined with expert experience and production rules. For example, the graph contains constraint relationships such as "machining center A must be processed before machining center B" and "product C needs to meet specific quality standards before entering the next process". The initial production capacity matrix data is checked and corrected for constraints using the knowledge graph reasoning engine. For example, if it is found that the production capacity of a certain process exceeds the processing capacity of subsequent processes, the production capacity of the process needs to be lowered according to the constraint relationship in the knowledge graph to avoid overcapacity and waste of resources. Based on the corrected production capacity matrix data, a comprehensive production capacity analysis is performed, such as calculating indicators such as overall production capacity, production efficiency, and production costs to generate comprehensive production capacity data.

[0087] Step S4: performing demand production forecasting based on the comprehensive production capacity data to obtain demand production forecasting data; performing order priority matching based on the demand production forecasting data to generate order priority allocation data;

[0088] In an embodiment of the present invention, the product demand in the future is predicted by combining historical order data, market demand forecast, inventory level and other information, for example, using time series analysis, machine learning and other methods to obtain demand production forecast data. For example, the demand for each product in the next month can be predicted based on the product sales data, seasonal laws, market trends and other information of the past year, and demand production forecast data can be generated. Then, based on the demand production forecast data, combined with the optimized production capacity matrix data, the urgency, importance, profit margin and other factors of different orders are comprehensively evaluated, for example, using weighted scoring method, hierarchical analysis method, etc., to generate order priority allocation data. For example, orders can be prioritized according to factors such as the delivery date of the order, customer type, product profit, etc., and urgent orders, important customer orders, and high-profit orders will be ranked at a higher priority.

[0089] Step S5: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data to generate preliminary subtask unit data; construct production subtask network data based on the preliminary subtask unit data; perform production plan conflict detection on the production subtask network data to obtain subtask production conflict data; use the subtask production conflict data to perform flexible scheduling processing on the production subtask network data to generate conflict-free production subtask network data;

[0090] In the embodiment of the present invention, the production process constraint knowledge graph is used to perform task decoupling processing on each order, decomposing it into multiple subtask units that can be executed in parallel or serially, such as processing, assembly, inspection, etc., to generate preliminary subtask unit data. According to the preliminary subtask unit data, combined with information such as equipment allocation and process flow sequence, the production subtask network data is constructed, for example, a directed acyclic graph (DAG) is used to represent the dependency relationship and execution order between subtasks. The constraint condition verification algorithm is used to analyze the production subtask network data, identify existing resource conflicts, time conflicts, logical conflicts, etc., such as multiple tasks need to occupy the same equipment at the same time, the completion time of the subtask exceeds the start time of the subsequent task, and the execution order of the subtask does not meet the process requirements, and the detection results are recorded in the subtask production conflict data. Flexible scheduling algorithms, such as genetic algorithms, simulated annealing algorithms, etc., are used to adjust and optimize the production subtask network data according to the subtask production conflict data, and the subtasks are re-decomposed or merged until conflict-free production subtask network data is generated.

[0091] Step S6: Identify the critical task path chain based on the conflict-free production subtask network data, and perform insertion planning of non-critical path tasks to generate task scheduling plan data; perform low-risk production plan execution processing on the task scheduling plan data to obtain the final production plan plan data.

[0092] In an embodiment of the present invention, a critical path algorithm is used to identify the critical task path chain that affects the overall production cycle, that is, the task sequence with the longest completion time. For example, among all the subtasks in the production of a product, "raw material preparation" requires 2 hours, "machining" requires 4 hours, "assembly" requires 3 hours, "testing" requires 1 hour, and "packaging" requires 0.5 hours. According to the critical path algorithm, "raw material preparation" → "machining" → "assembly" constitutes the critical path, because this path takes the longest total time, which determines the completion time of the entire order. For tasks on non-critical paths, an insert planning algorithm is used to flexibly arrange them in combination with information such as the available time window of the equipment and the skill level of the personnel. Based on the task arrangement results of the critical path and the non-critical path, task scheduling plan data is generated, which records in detail the start time, end time, execution equipment, person in charge and other information of each subtask.

[0093] Preferably, step S1 comprises the following steps:

[0094] Step S11: Use the industrial Internet data platform to collect real-time data from the production site to obtain the original industrial data stream;

[0095] Step S12: performing collaborative fusion of heterogeneous data according to the original industrial data stream to generate fused original industrial data;

[0096] Step S13: performing adaptive data cleaning and filtering on the fused original industrial data to obtain cleaned industrial production data;

[0097] Step S14: performing timestamp calibration processing based on the original industrial production data to generate industrial production calibration data;

[0098] Step S15: Perform data association aggregation processing according to the industrial production calibration data, and perform industrial production context processing to generate real-time industrial production context data.

[0099] In an embodiment of the present invention, the industrial Internet data platform collects various data in the production process in real time through various sensors, RFID tags, numerical control systems and other equipment deployed at the production site, such as equipment operating status, temperature, pressure, flow, material consumption, product quality detection data, etc. These data are transmitted to the data platform through industrial field buses, wireless networks, etc., forming a multi-source heterogeneous original industrial data stream. Data mapping technology is used to convert data in different formats into a unified data format, such as converting data of different protocols into OPCUA format, and converting data of different units into a unified unit. Data alignment technology is used to associate and merge data from different data sources to generate fused original industrial data, such as merging data from temperature sensors, pressure sensors, and flow meters in the same time period to form production process data in the time period. Rule-based methods, such as setting thresholds, logical judgments, etc., are used to identify and process obvious outliers and erroneous data, such as identifying data exceeding the sensor range as outliers and identifying data with incorrect status codes as erroneous data. Based on the data timestamp and data transmission delay, etc., time synchronization algorithms such as NTP protocol and PTP protocol are used to align the data from different data sources, for example, calibrate the timestamps of all data to the system time of the data platform. According to the pre-defined production process and equipment association relationship, the associated and aggregated data is analyzed and processed. For example, the processing time, process parameters, quality inspection results and other information of the product in different processes and different equipment are tracked according to the product ID to form the real-time production context data of the product.

[0100] Preferably, step S2 comprises the following steps:

[0101] Step S21: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data, wherein the real-time industrial production resource data includes industrial production equipment node data and staff node data;

[0102] Step S22: dynamically updating equipment parameters of a preset production equipment digital model through industrial production equipment node data to generate a dynamic production equipment digital model;

[0103] Step S23: quantifying the equipment production capacity of the dynamic production equipment digital model to obtain real-time equipment production capacity data;

[0104] Step S24: extracting personnel skills according to the personnel node data, and matching personnel skills through a preset personnel skill model library to generate personnel skill data; performing production skill proficiency assessment processing on the personnel skill data through the personnel node data to generate personnel production capacity data;

[0105] Step S25: performing production capacity processing on the real-time industrial production resource data based on the personnel production capacity data and the real-time equipment production capacity data to obtain a production capacity dynamic index;

[0106] Step S26: label the real-time industrial production context data with production capacity through the production capacity dynamic index, and construct a production capacity matrix to obtain initial production capacity matrix data.

[0107] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S2 in the embodiment are shown in the flowchart. In this embodiment, step S2 includes:

[0108] Step S21: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data, wherein the real-time industrial production resource data includes industrial production equipment node data and staff node data;

[0109] In an embodiment of the present invention, the real-time industrial production context data includes detailed data of each link in the production process, such as equipment operation status, material consumption, product quality, etc. Information related to production resources, such as equipment ID, equipment type, personnel ID, working hours, etc., is extracted from the real-time industrial production context data. Combined with the pre-configured production resource information library in the data platform, such as the equipment information table, the personnel information table, etc., real-time industrial production resource data can be generated. The data includes two parts: industrial production equipment node data and staff node data. For example, it is extracted that the product uses A equipment and B equipment in the processing link, which are operated by operators Zhang San and Li Si respectively. Combined with the equipment information table and the personnel information table, industrial production equipment node data containing information such as the operation status of equipment A and equipment B, parameter configuration, etc., and staff node data containing information such as the working hours and skill levels of Zhang San and Li Si can be generated.

[0110] Step S22: dynamically updating equipment parameters of a preset production equipment digital model through industrial production equipment node data to generate a dynamic production equipment digital model;

[0111] In the embodiment of the present invention, the digital models of all production equipment are pre-stored in the industrial Internet platform. These models contain static data such as geometric information, physical characteristics, and operating parameters of the equipment. The real-time parameters in the industrial production equipment node data obtained in step S21, such as temperature, pressure, speed, etc., can be updated to the corresponding equipment digital model. Parameter update can be achieved by data mapping, for example, mapping the "A device temperature" field in the device node data to the "temperature" attribute in the A device digital model.

[0112] Step S23: quantifying the equipment production capacity of the dynamic production equipment digital model to obtain real-time equipment production capacity data;

[0113] In the embodiment of the present invention, the dynamic production equipment digital model includes the real-time parameters and operating status of the equipment, and the production capacity of the equipment is quantitatively evaluated according to a preset equipment capacity evaluation model, such as a method based on a physical model, a method based on data-driven, etc., for example, the equipment's unit time output, qualified rate, energy consumption and other indicators are calculated. Different types of equipment can use different evaluation models. For example, for CNC machine tools, a model based on cutting force can be used to evaluate their processing capabilities; for robotic arms, a model based on kinematics and dynamics can be used to evaluate their operating capabilities.

[0114] Step S24: extracting personnel skills according to the personnel node data, and matching personnel skills through a preset personnel skill model library to generate personnel skill data; performing production skill proficiency assessment processing on the personnel skill data through the personnel node data to generate personnel production capacity data;

[0115] In an embodiment of the present invention, the staff node data includes information such as the staff's working hours and skill level. The specific skills of the staff, such as "CNC machine tool operation" and "product quality inspection", can be extracted from the staff node data according to preset skill extraction rules, such as regular expressions and keyword matching. Using the staff skill model library, the extracted skills can be matched with the skills in the model library, for example, the "CNC machine tool operation" skill can be matched to the "CNC machine tool operation" skill model. Based on the working hours, operation records and other information in the staff node data, the skill proficiency of the personnel is evaluated, for example, the personnel's operation proficiency, product qualification rate and other indicators are calculated.

[0116] Step S25: performing production capacity processing on the real-time industrial production resource data based on the personnel production capacity data and the real-time equipment production capacity data to obtain a production capacity dynamic index;

[0117] In the embodiment of the present invention, the real-time industrial production resource data can be comprehensively analyzed to calculate the production capacity dynamic index. The calculation method of the production capacity dynamic index can be designed according to the specific application scenario. For example, the production capacity of the equipment, the efficiency of the personnel, the utilization rate of the resources and other factors can be considered, and the weighted average, fuzzy comprehensive evaluation and other methods can be used for calculation.

[0118] Step S26: label the real-time industrial production context data with production capacity through the production capacity dynamic index, and construct a production capacity matrix to obtain initial production capacity matrix data.

[0119] In an embodiment of the present invention, the index is annotated to the real-time industrial production context data, for example, the production capacity dynamic index of each production link is added to the corresponding process data. According to the real-time industrial production context data annotated with the production capacity dynamic index, the initial production capacity matrix data is constructed. The matrix reflects the production capacity differences of different production links under the current resource configuration and production status. For example, each element in the matrix represents the comprehensive production capacity index of a specific process in a specific time period.

[0120] Preferably, step S25 comprises the following steps:

[0121] Step S251: Evaluate the equipment load through the industrial production equipment node data to generate production equipment load index data;

[0122] Step S252: Perform equipment availability analysis based on production equipment load index data to generate an equipment availability index;

[0123] Step S253: Performing personnel availability analysis through the personnel node data to generate a personnel availability index;

[0124] Step S254: using the equipment availability index and the personnel availability index to perform spatiotemporal production node status evaluation on the real-time industrial production context data to generate spatiotemporal production status evaluation data;

[0125] Step S255: Perform productivity field processing according to the spatiotemporal production status evaluation data to obtain industrial productivity distribution data;

[0126] Step S256: Perform capacity decay analysis on the industrial productivity field data based on the personnel production capacity data and the real-time equipment production capacity data to generate capacity decay data;

[0127] Step S257: performing capacity compensation processing according to the capacity decay data to obtain productivity capacity compensation data;

[0128] Step S258: Calculate the production capacity dynamic index of the industrial productivity distribution data using the productivity capacity compensation data to generate the production capacity dynamic index.

[0129] In the embodiment of the present invention, the industrial production equipment node data includes the real-time operation status, processing parameters, energy consumption and other information of the equipment. For example, the comprehensive load index of the equipment can be calculated based on the real-time CPU usage rate, memory occupancy rate, disk IO and other indicators of the equipment, and the index reflects the busyness of the equipment in the current time period. In combination with a preset equipment availability analysis model, such as a threshold-based model, a machine learning-based model, etc., the availability of the equipment can be analyzed. For example, the probability of failure of the equipment in the future can be predicted based on the load index change trend of the equipment, historical fault records and other information, and the availability index of the equipment can be calculated based on the failure probability. According to preset rules or models, such as work time schedules, personnel skill matrices, etc., the availability of personnel is analyzed. For example, the probability that the personnel can participate in production in the future can be calculated based on the current working status of the personnel, skill matching, historical leave records and other information, and the availability index of the personnel can be calculated based on the probability. For example, based on the equipment availability index of a process, the availability index of the operator, the position of the process in the production process and other information, it is possible to evaluate whether the process can be carried out normally in the future, and calculate the spatiotemporal production status evaluation value of the process. Each process in the production system is regarded as a particle with "productivity", and the interaction force between the particles is determined by the spatiotemporal production state evaluation value. By calculating the distribution of the force field, the industrial productivity distribution data can be obtained, which reflects the distribution of production capacity in time and space. For example, extract the average output per worker per hour, the qualified rate and other indicators, and compare them with the baseline level of the worker. If the indicator is lower than the baseline level, it means that the production capacity of the worker has declined. For example, extract the real-time operating parameters of the equipment (such as temperature, pressure, vibration, etc.) and compare them with the preset normal range. If the parameter exceeds the normal range, it means that the production capacity of the equipment has declined. Map the decline of personnel and equipment to the industrial productivity field data. For example, a process is operated by a worker in a specific time period, and the worker's production capacity declines by 10%, then the predicted value of the production capacity of the process in the time period is reduced by 10%. For example, according to the cause of capacity decline, measures such as increasing manpower, optimizing shift scheduling, adjusting equipment parameters, and replacing raw materials can be taken. For example, historical data analysis, simulation experiments and other methods can be used to determine how much production can be increased by adding one person, how much production cycle can be shortened by optimizing shift scheduling, etc. According to the capacity decay data and the effect of capacity compensation measures, the industrial productivity field data is corrected. For example, if the production capacity of a process decays by 10% in a specific time period, and adding one more person can increase the production capacity of the process by 15%, then the production capacity forecast value of the process in the time period is increased by 5%. The productivity capacity compensation data is integrated with the industrial productivity distribution data.Specifically, the capacity compensation value corresponding to each data point can be added to the corresponding industrial productivity distribution data point. For example, if the industrial productivity distribution data value of a process in a specific time period is 100, and the capacity compensation value in this time period is 5, then the fused value is 105. The production capacity dynamic index of each process in each time period is calculated based on the fused data.

[0130] Preferably, step S3 comprises the following steps:

[0131] Step S31: Identify production process elements on real-time industrial production context data to generate production process element data;

[0132] Step S32: extracting constraints based on the production process element data to generate constraint structure data, wherein the constraint structure data includes production process flow, quality control standards and safety specifications; constructing a production process constraint knowledge graph based on the constraint structure data;

[0133] Step S33: Mapping the initial production capacity matrix data to production constraints through the production process constraint knowledge graph to generate production constraint mapping data;

[0134] Step S34: performing constraint condition association analysis on the production constraint mapping data, and performing constraint condition priority sorting to generate constraint priority data;

[0135] Step S35: dynamically adjusting the constraint conditions of the production constraint mapping data through the constraint priority data to obtain actual production capacity constraint data;

[0136] Step S36: using the actual production capacity constraint data to perform actual capacity value correction processing on the initial production capacity matrix data to obtain optimized production capacity matrix data;

[0137] Step S37: Calculate the comprehensive production capacity based on the optimized production capacity matrix data to generate comprehensive production capacity data.

[0138] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S3 in the flowchart, in this example, step S3 includes:

[0139] Step S31: Identify production process elements on real-time industrial production context data to generate production process element data;

[0140] In an embodiment of the present invention, real-time industrial production context data is parsed and extracted. For example, regular expressions can be used to match all records involving equipment operations, and extract information such as equipment ID, operation type, and operation time; natural language processing technology can be used to identify all records involving quality inspections, and extract information such as inspection items, inspection results, and judgment criteria. All extracted element information is integrated to form production process element data. This data contains key elements that have a constraining effect on the production process, such as equipment, personnel, materials, environment, process parameters, quality indicators, etc.

[0141] Step S32: extracting constraints based on the production process element data to generate constraint structure data, wherein the constraint structure data includes production process flow, quality control standards and safety specifications; constructing a production process constraint knowledge graph based on the constraint structure data;

[0142] In an embodiment of the present invention, constraint structure data is extracted based on production process element data. For example, based on the equipment operation record, the constraint condition "equipment A must operate in an environment with a temperature below 50°C" can be extracted; based on the quality inspection record, the constraint condition "the diameter of product B must be within the range of 10±0.5mm" can be extracted. Constraint subject: refers to the constrained object, such as equipment, products, process parameters, etc. Constraint conditions: refers to specific requirements for the constraint subject, such as numerical range, operating procedures, time limits, etc. Constraint source: refers to the basis of the constraint conditions, such as design specifications, safety standards, operating manuals, etc. The extracted constraint structure data, as well as information such as production process flow, quality control standards, safety specifications, etc., are imported into the knowledge graph construction tool. In the knowledge graph, each node represents an entity, such as equipment, product, process parameter, constraint condition, etc.; each edge represents the relationship between entities, such as "operation", "control", "constraint", etc.

[0143] Step S33: Mapping the initial production capacity matrix data to production constraints through the production process constraint knowledge graph to generate production constraint mapping data;

[0144] In an embodiment of the present invention, the initial production capacity matrix data reflects the production capacity of each production link under ideal conditions, while in the actual production process, each link needs to meet certain constraints. By mapping the constraints to the corresponding production links, it is possible to identify which links are subject to which constraints. For example, assuming that the initial production capacity matrix data shows that the production capacity of process A is 100 pieces / hour. Through the query of the production process constraint knowledge graph, it is found that process A is subject to the constraint that "equipment A must operate in an environment with a temperature below 50°C". Then, the constraint is mapped to process A to generate production constraint mapping data.

[0145] Step S34: performing constraint condition association analysis on the production constraint mapping data, and performing constraint condition priority sorting to generate constraint priority data;

[0146] In the embodiment of the present invention, the priority of the constraint condition can be evaluated according to factors such as its importance, urgency, and scope of influence. The priority of the constraint condition can be evaluated according to factors such as its importance, urgency, and scope of influence. The priority scores of all the constraints are sorted to generate constraint priority data. The constraints include: safety-related constraints, quality-related constraints, and efficiency-related constraints; the priorities include: highest, higher, and general.

[0147] Step S35: dynamically adjusting the constraint conditions of the production constraint mapping data through the constraint priority data to obtain actual production capacity constraint data;

[0148] In an embodiment of the present invention, for example, a certain process is subject to three constraints at the same time, namely, a safety-related constraint, a quality-related constraint, and an efficiency-related constraint. According to the constraint priority data, the safety-related constraint has the highest priority, so it is necessary to give priority to satisfying this constraint. If this constraint causes the production capacity of the process to be reduced by 20%, then the actual production capacity constraint data of the process is set to be reduced by 20%. On the premise of satisfying the safety-related constraint, the quality-related constraint can also be satisfied, then the actual production capacity constraint data of the process is continued to be reduced until all constraints cannot be satisfied at the same time. The actual production capacity constraint data obtained reflects the maximum production capacity limit that can actually be achieved in each production link when all high-priority constraints are satisfied.

[0149] Step S36: using the actual production capacity constraint data to perform actual capacity value correction processing on the initial production capacity matrix data to obtain optimized production capacity matrix data;

[0150] In the embodiment of the present invention, for example, assuming that the initial production capacity matrix data shows that the production capacity of process A is 100 pieces / hour, and the actual production capacity constraint data shows that the actual production capacity limit of process A is 80 pieces / hour. Then, the production capacity value of process A in the optimized production capacity matrix data is corrected to 80 pieces / hour. By correcting all processes, the optimized production capacity matrix data finally obtained reflects the maximum production capacity that can actually be achieved in each production link when all constraints are met.

[0151] Step S37: Calculate the comprehensive production capacity based on the optimized production capacity matrix data to generate comprehensive production capacity data.

[0152] In the embodiment of the present invention, for example, bottleneck analysis method, linear programming method and other methods can be used to calculate the maximum number of products that can be produced by the entire production system in a unit time and the production ratio of different products based on the production capacity, production rhythm, product process route and other information of each process. The comprehensive production capacity data finally obtained reflects the maximum production capacity that the entire production system can actually achieve when all constraints are met.

[0153] Preferably, step S4 comprises the following steps:

[0154] Step S41: Acquire historical production order data and real-time production order data; perform feature engineering processing on the historical production order data to generate historical production order feature data;

[0155] Step S42: performing demand production forecasting on the historical production order feature data by using a preset autoregressive difference moving average model to obtain demand production forecasting data;

[0156] Step S43: using the comprehensive production capacity data to dynamically adjust the demand production forecast data to generate revised demand production forecast data;

[0157] Step S44: performing order urgency processing on the real-time production order data to generate order production urgency data;

[0158] Step S45: Calculate the order priority of the order production urgency data by modifying the demand production forecast data to generate order priority allocation data.

[0159] In an embodiment of the present invention, the industrial Internet platform stores a large amount of historical production order data, which records the detailed information of past orders, such as product type, quantity, delivery date, customer information, etc. At the same time, the platform also receives new production order data in real time. The historical production order data is cleaned and preprocessed, such as processing missing values, outliers, duplicate values, etc., and the data format is unified. The autoregressive differential moving average model (ARIMA) is a commonly used time series prediction model, which can predict future trends based on the laws of time series data. Using the historical data of the industrial Internet platform, the ARIMA model can be trained, and the historical production order feature data can be input to predict the production demand for a period of time in the future, such as the demand for each product in the next week. The comprehensive production capacity data reflects the maximum production capacity of each product in each time period under the current resource configuration and production conditions. For example, assuming that the demand production forecast data shows that the demand for product A in the next week is 1,000 pieces, and the comprehensive production capacity data shows that the maximum production capacity of product A in the next week is 800 pieces. Then, it is necessary to revise the demand forecast value of product A and adjust it to 800 pieces to ensure the feasibility of the production plan. The urgency of an order can be assessed based on factors such as delivery date, customer priority, and product importance. For example, a rule engine can be used to rate the urgency of orders based on pre-set rules. For example: Orders with a delivery date within 3 days are rated "urgent". Orders from VIP customers are rated "important". Orders involving critical products are rated "urgent". Priority is calculated for all orders and the final order priority allocation data is generated. For example, a weighted scoring method can be used to calculate the priority score of each order based on the predicted demand and urgency of the order. The priority scores of all orders are sorted to generate order priority allocation data.

[0160] Preferably, step S5 comprises the following steps:

[0161] Step S51: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data, and perform subtask process module mapping to generate preliminary subtask unit data;

[0162] Step S52: clustering the demand clusters according to the preliminary subtask unit data, and constructing the task network to obtain the production subtask network data;

[0163] Step S53: Prioritize the production subtask network data using the order priority allocation data to obtain priority production subtask network data;

[0164] Step S54: Performing production plan conflict detection on the priority production subtask network data by optimizing the production capacity matrix data and the production process constraint knowledge graph to obtain subtask production conflict data, wherein the subtask production conflict data includes subtask time conflict data, subtask production resource conflict data and subtask production constraint violation data;

[0165] Step S55: Utilize the subtask production conflict data to perform flexible scheduling processing on the priority production subtask network data to generate conflict-free production subtask network data.

[0166] In an embodiment of the present invention, the production process flow of the product is queried in the production process constraint knowledge graph according to the product type, process route and other information of the order. For example, assuming that order A needs to produce product X, the production process flow of product X can be queried as follows: raw material preparation → processing process 1 → processing process 2 → assembly process → inspection process → packaging process. According to pre-defined rules or models, each production process is decomposed into more fine-grained sub-task units and mapped with the process modules in the production process constraint knowledge graph. For example, "processing process 1" can be decomposed into three sub-task units of "loading", "processing" and "unloading", and these three sub-task units are respectively mapped to the corresponding "loading module", "processing module" and "unloading module" in the knowledge graph. According to the process module, input and output and other information of each sub-task unit, as well as the process flow information in the production process constraint knowledge graph, the dependency relationship and execution order between each sub-task unit are determined. For example, assuming that the output of sub-task unit A is the input of sub-task unit B, then A and B are connected, and A is set as the predecessor task of B. All subtask units and their dependencies are represented in the form of a directed acyclic graph (DAG) to generate production subtask network data. According to the order to which each subtask unit belongs, the priority of the order is obtained from the order priority allocation data and marked on the corresponding subtask unit. For example, assuming that the priority of order A is "high", then the priority of all subtask units belonging to order A is marked as "high". According to the optimized production capacity matrix data, the availability of each resource in each time period is obtained. For example, assuming that device A has been occupied between 8:00-9:00, then the status of device A in this time period is set to "unavailable". According to the production process constraint knowledge graph, the constraints of each subtask unit are obtained. For example, assuming that subtask unit A must be completed on device B, then device B is set as the constraint of A. For example, if there is a resource conflict between two subtask units, and one of the subtask units has a higher priority, then the resource requirements of the high-priority subtask unit are met first, and the low-priority subtask unit is scheduled to be executed in other time periods or on other resources. By iteratively adjusting the priority production subtask network data for multiple times until all conflicts are resolved, conflict-free production subtask network data is finally generated. This data contains information such as the final execution time, execution resources, and execution order of all subtask units.

[0167] Preferably, step S54 includes the following steps:

[0168] Step S541: Perform production time series analysis on the priority production subtask network data to obtain production subtask time series data;

[0169] Step S542: Perform time constraint processing according to the production subtask timing data, and perform subtask time conflict detection to generate subtask time conflict data;

[0170] Step S543: Perform production resource demand analysis on the priority production subtask network data to generate production resource demand data;

[0171] Step S544: matching the production resource demand data with the production resource capacity by optimizing the production capacity matrix data to generate resource capacity matching data;

[0172] Step S545: Perform production resource conflict detection according to the resource capability matching data to obtain subtask production resource conflict data;

[0173] Step S546: Perform production process constraint violation detection on the priority production subtask network data through the production process constraint knowledge graph to obtain subtask production constraint violation data.

[0174] In the embodiment of the present invention, according to the estimated execution time of each subtask and the dependency relationship between the subtasks, the earliest start time, the latest start time, the earliest end time, the latest end time and other time parameters of each subtask are calculated using algorithms such as the critical path method (CPM). The calculated time parameters are integrated with other information of the subtask (such as subtask ID, order, execution process, etc.) to generate production subtask timing data. For example, suppose that the earliest start time of subtask A is 8:00 and the latest end time is 9:00, while the earliest start time of subtask B is 8:30 and the latest end time is 9:30. Since the execution time periods of A and B overlap, there is a potential time conflict between them. By traversing all subtask pairs and comparing their earliest start time, latest end time and other parameters, all subtask pairs with time conflicts can be identified. These conflict information are integrated to generate subtask time conflict data. For example, suppose that subtask A needs to be processed using A equipment, needs to be operated by operator Zhang San, and needs to consume 10kg of raw material X. Then record the A device, Zhang San, and raw material X as the resource requirements of subtask A. Integrate the resource requirement information of all subtasks to generate production resource requirement data. The optimized production capacity matrix data reflects the available capacity of each resource in each time period. For example, the data can show that the A device can process 100 products between 8:00-9:00. Compare the resource requirements of each subtask with the available capacity of the resources in the corresponding time period. If the demand exceeds the available capacity, it means that the resource requirements of the subtask cannot be met. For example, suppose that both subtask A and subtask B need to use the A device between 8:00-9:00, and their resource requirements cannot be met, then mark A and B as having a resource conflict. Integrate all subtask pairs with resource conflicts to generate subtask production resource conflict data. For example, suppose that the production process constraint knowledge graph stipulates that the processing steps of product X must be completed on the A device and the B device in sequence, and in the priority production subtask network data, the processing subtask of product X is scheduled to be executed on the C device, then it is considered that the subtask violates the production process constraint. By traversing all subtasks and comparing their execution methods with the constraints in the production process constraint knowledge graph, all subtasks with constraint violations can be identified. These violation information are integrated to generate subtask production constraint violation data.

[0175] Preferably, step S55 comprises the following steps:

[0176] Step S551: performing time constraint relaxation processing on the subtask time conflict data to generate subtask production relaxation time data;

[0177] Step S552: rescheduling the priority production subtask network data according to the subtask production slack time data, and screening the optimal time conflict scheduling according to the preset time flexibility priority adjustment rule to generate the optimal time conflict scheduling data;

[0178] Step S553: ​​performing alternative production resource analysis on the subtask production resource conflict data to generate alternative production resource data;

[0179] Step S554: rescheduling the network data of the priority production subtasks to resolve resource conflicts through the replaceable production resource data, and adjusting the load balancing optimization to generate optimal resource conflict allocation data;

[0180] Step S555: performing constraint relaxation determination on the subtask production constraint violation data, and generating loosenable constraint data;

[0181] Step S556: adjusting the constraint parameters of the priority production subtask network data through the releasable constraint data, and performing production constraint risk assessment to generate low-risk constraint adjustment data;

[0182] Step S557: Use low-risk constraint adjustment data, optimal resource conflict allocation data, and optimal time conflict scheduling data to eliminate conflicts in the priority production subtask network data, perform global conflict verification, and generate conflict-free production subtask network data.

[0183] In an embodiment of the present invention, the earliest start time, the latest start time and other information of each subtask in the production subtask time series data, as well as the preset slack time limit, are combined to calculate the time range that each subtask can adjust, and generate the subtask production slack time data. According to the subtask production slack time data, the execution time of the conflicting subtask is adjusted, for example, the start time of subtask A is adjusted to 7:45, and the start time of subtask B is adjusted to 8:45, so as to resolve the time conflict. Since time adjustment will generate new conflicts, multiple rounds of iterative adjustment are required until all time conflicts are resolved. In each round of iteration, multiple feasible scheduling schemes can be evaluated and screened according to the preset time flexibility priority adjustment rules, such as "prioritize the time of low-priority subtasks", "prioritize the subtasks with a larger slack time range", etc., and finally the optimal time conflict scheduling scheme is selected. For example, suppose that both subtask A and subtask B need to use device A, resulting in resource conflict. By querying the optimized production capacity matrix data and the production process constraint knowledge map, it is found that device B is available within the time period and can be used to complete the processing of subtask A. Then device B is recorded as an alternative resource for device A. For example, suppose subtask A violates the constraint condition of "must be completed on device A". By querying the knowledge graph of production process constraints, it is found that the priority of this constraint condition is "suggested" and there is an alternative solution of "can be completed on device B". Then the constraint condition is marked as "relaxable". For example, the execution device of subtask A is adjusted from device A to device B to meet the relaxed constraint condition. When adjusting constraint parameters, it is necessary to evaluate the risks brought by the adjustment, such as: quality risk: whether the adjusted solution will affect product quality, efficiency risk: whether the adjusted solution will reduce production efficiency, safety risk: whether the adjusted solution will bring safety hazards. After completing all adjustments, it is necessary to perform global conflict verification on the entire production plan to ensure that there are no new conflicts. For example, it is necessary to check whether the adjusted subtask execution time is still within the available time range of the resources, whether the adjusted subtask execution resources still meet the constraints, etc. If new conflicts are found during the verification process, it is necessary to return to the previous steps and readjust them until all conflicts are resolved.

[0184] Preferably, step S6 comprises the following steps:

[0185] Step S61: identifying the critical task path chain according to the conflict-free production subtask network data, and generating critical path task chain data;

[0186] Step S62: Performing operation time window processing according to the critical path task chain data to generate critical operation path timing data;

[0187] Step S63: Perform non-critical path task insertion planning on the conflict-free production subtask network data through the critical operation path timing data to generate task scheduling plan data;

[0188] Step S64: Perform a scheduling risk assessment on the task scheduling data, and perform low-risk production plan execution processing to obtain final production plan data.

[0189] In an embodiment of the present invention, the critical path method (CPM) is used to analyze the conflict-free production subtask network data. The method regards each subtask as a node, regards the dependency relationship between subtasks as a directed edge, and calculates the earliest start time, the latest start time, the earliest end time, and the latest end time of each node according to the execution time of each subtask. By comparing all the paths from the starting point to the end point, the path with the longest total execution time is found, which is the critical path. All subtasks on the critical path are extracted and arranged in order of execution to generate critical path task chain data. According to the execution time and the previous and next dependencies of each subtask in the critical path task chain data, the available time window of each subtask is calculated. For example, assuming that subtask A is a task on the critical path, its execution time is 8:00-9:00, the latest end time of its predecessor task B is 7:30, and the earliest start time of its successor task C is 9:30, then the available time window of A is 7:30-9:30. According to the resource requirements, time constraints and other information of each non-critical path task, as well as the available resources, available time and other information of each time window in the critical operation path timing data, the non-critical path tasks are inserted into the appropriate time window using methods such as greedy algorithms and heuristic algorithms. Perform a comprehensive analysis of the task scheduling data to identify potential risk points. The analysis process can be aided by risk assessment models and historical data accumulated by the industrial Internet platform. For example: resource availability risk analysis, equipment failure risk analysis, and personnel operation risk analysis. After completing the risk assessment, a risk assessment report containing information such as various risk points, risk levels, and risk occurrence probabilities is generated. According to the risk assessment report, the task scheduling data is adjusted to reduce risks by adjusting the start and end times of the tasks, reallocating resources, or taking emergency measures (such as preparing additional raw materials, spare equipment, or rapid response teams). On the premise of ensuring that risks are controllable, the optimized task scheduling plan is converted into the final production plan data.

[0190] The beneficial effect of this application is that the real-time collection of production site data through the industrial Internet data platform ensures the timeliness and accuracy of the data. The industrial production context processing not only integrates the original data, but also reveals the internal connections and flow patterns in the production process, which helps to deeply understand the production status and process. Through resource node analysis and production capacity analysis, real-time industrial production resource data and dynamic production capacity index are generated, which can quickly locate resource bottlenecks, reflect production capacity changes in real time, build production capacity matrix and production process constraint knowledge map, comprehensively depict the capacity state and constraint relationship of the production system, and enhance the understanding and control ability of complex production environments. The production capacity matrix is ​​corrected by using the knowledge map to ensure the feasibility and rationality of the production plan, reflect the optimal production capacity layout under all actual constraints, and can more accurately estimate future production demand, providing forward-looking guidance for resource allocation and production scheduling. The production sequence is reasonably arranged according to the forecast results to ensure that high-priority orders are completed in time, which improves customer satisfaction and market response speed. Task decoupling processing splits complex orders into executable subtask units, refines and optimizes the production task structure. The construction of the production subtask network and the conflict detection mechanism effectively avoid time and resource conflicts. The identification of key task paths optimizes the task execution sequence, maximizes the utilization of production resources, and improves production efficiency. The low-risk production plan execution process ensures the stable implementation of the final production plan. It comprehensively considers efficiency, cost and risk, has a high degree of executableness, realizes intelligent and refined production management, and improves production efficiency.

[0191] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0192] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A data processing method based on an industrial Internet data platform, characterized in that: The following steps are involved: Step S1: Use the industrial Internet data platform to collect real-time data on the production site to obtain the original industrial data stream; perform industrial production context processing based on the original industrial data stream to generate real-time industrial production context data; Step S2: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data; perform production capacity analysis on the real-time industrial production resource data to obtain a production capacity dynamic index; construct a production capacity matrix for the real-time industrial production context data using the production capacity dynamic index to obtain initial production capacity matrix data; Step S3: construct a production process constraint knowledge graph based on real-time industrial production context data; The initial production capacity matrix data is processed by production constraint correction using the production process constraint knowledge graph to generate optimized production capacity matrix data; comprehensive production capacity analysis is performed based on the optimized production capacity matrix data to generate comprehensive production capacity data; wherein step S3 is specifically as follows: Step S31: Identify production process elements on real-time industrial production context data to generate production process element data; Step S32: extracting constraints based on the production process element data to generate constraint structure data, wherein the constraint structure data includes production process flow, quality control standards and safety specifications; constructing a production process constraint knowledge graph based on the constraint structure data; Step S33: Mapping the initial production capacity matrix data to production constraints through the production process constraint knowledge graph to generate production constraint mapping data; Step S34: performing constraint condition association analysis on the production constraint mapping data, and performing constraint condition priority sorting to generate constraint priority data; Step S35: dynamically adjusting the constraint conditions of the production constraint mapping data through the constraint priority data to obtain actual production capacity constraint data; Step S36: using the actual production capacity constraint data to perform actual capacity value correction processing on the initial production capacity matrix data to obtain optimized production capacity matrix data; Step S37: Calculate the comprehensive production capacity according to the optimized production capacity matrix data to generate comprehensive production capacity data; Step S4: performing demand production forecasting based on the comprehensive production capacity data to obtain demand production forecasting data; performing order priority matching based on the demand production forecasting data to generate order priority allocation data; Step S5: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data to generate preliminary subtask unit data; construct production subtask network data based on the preliminary subtask unit data; perform production plan conflict detection on the production subtask network data to obtain subtask production conflict data; use the subtask production conflict data to perform flexible scheduling processing on the production subtask network data to generate conflict-free production subtask network data; Step S6: Identify the critical task path chain according to the conflict-free production subtask network data, and perform non-critical path task insertion planning to generate task scheduling plan data; perform low-risk production plan execution processing on the task scheduling plan data to obtain the final production plan plan data; wherein, step S6 is specifically as follows: Step S61: identifying the critical task path chain according to the conflict-free production subtask network data, and generating critical path task chain data; Step S62: Performing operation time window processing according to the critical path task chain data to generate critical operation path timing data; Step S63: Perform non-critical path task insertion planning on the conflict-free production subtask network data through the critical operation path timing data to generate task scheduling plan data; the non-critical path task insertion planning is to insert the non-critical path task into a suitable time window; Step S64: Perform a scheduling risk assessment on the task scheduling data, and perform low-risk production plan execution processing to obtain final production plan data.

2. The data processing method based on the industrial Internet data platform according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Use the industrial Internet data platform to collect real-time data from the production site to obtain the original industrial data stream; Step S12: performing collaborative fusion of heterogeneous data according to the original industrial data stream to generate fused original industrial data; Step S13: performing adaptive data cleaning and filtering on the fused original industrial data to obtain cleaned industrial production data; Step S14: performing timestamp calibration processing based on the original industrial production data to generate industrial production calibration data; Step S15: Perform data association aggregation processing according to the industrial production calibration data, and perform industrial production context processing to generate real-time industrial production context data.

3. The data processing method based on the industrial Internet data platform according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: Perform production resource node analysis based on real-time industrial production context data to generate real-time industrial production resource data, wherein the real-time industrial production resource data includes industrial production equipment node data and staff node data; Step S22: dynamically updating equipment parameters of a preset production equipment digital model through industrial production equipment node data to generate a dynamic production equipment digital model; Step S23: quantifying the equipment production capacity of the dynamic production equipment digital model to obtain real-time equipment production capacity data; Step S24: extracting personnel skills according to the personnel node data, and matching personnel skills through a preset personnel skill model library to generate personnel skill data; performing production skill proficiency assessment processing on the personnel skill data through the personnel node data to generate personnel production capacity data; Step S25: performing production capacity processing on the real-time industrial production resource data based on the personnel production capacity data and the real-time equipment production capacity data to obtain a production capacity dynamic index; Step S26: label the real-time industrial production context data with production capacity through the production capacity dynamic index, and construct a production capacity matrix to obtain initial production capacity matrix data.

4. The data processing method based on the industrial Internet data platform according to claim 3 is characterized in that: Step S25 includes the following steps: Step S251: Evaluate the equipment load through the industrial production equipment node data to generate production equipment load index data; Step S252: Perform equipment availability analysis based on production equipment load index data to generate an equipment availability index; Step S253: Performing personnel availability analysis through the personnel node data to generate a personnel availability index; Step S254: using the equipment availability index and the personnel availability index to perform spatiotemporal production node status evaluation on the real-time industrial production context data to generate spatiotemporal production status evaluation data; Step S255: Perform productivity field processing according to the spatiotemporal production status evaluation data to obtain industrial productivity distribution data; Step S256: Perform capacity decay analysis on the industrial productivity field data based on the personnel production capacity data and the real-time equipment production capacity data to generate capacity decay data; Step S257: performing capacity compensation processing according to the capacity decay data to obtain productivity capacity compensation data; Step S258: Calculate the production capacity dynamic index of the industrial productivity distribution data using the productivity capacity compensation data to generate the production capacity dynamic index.

5. The data processing method based on the industrial Internet data platform according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: Acquire historical production order data and real-time production order data; perform feature engineering processing on the historical production order data to generate historical production order feature data; Step S42: performing demand production forecasting on the historical production order feature data by using a preset autoregressive difference moving average model to obtain demand production forecasting data; Step S43: using the comprehensive production capacity data to dynamically adjust the demand production forecast data to generate revised demand production forecast data; Step S44: performing order urgency processing on the real-time production order data to generate order production urgency data; Step S45: Calculate the order priority of the order production urgency data by modifying the demand production forecast data to generate order priority allocation data.

6. The data processing method based on the industrial Internet data platform according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: Use the production process constraint knowledge graph to perform task decoupling processing on the order priority allocation data, and perform subtask process module mapping to generate preliminary subtask unit data; Step S52: clustering the demand clusters according to the preliminary subtask unit data, and constructing the task network to obtain the production subtask network data; Step S53: Prioritize the production subtask network data using the order priority allocation data to obtain priority production subtask network data; Step S54: Performing production plan conflict detection on the priority production subtask network data by optimizing the production capacity matrix data and the production process constraint knowledge graph to obtain subtask production conflict data, wherein the subtask production conflict data includes subtask time conflict data, subtask production resource conflict data and subtask production constraint violation data; Step S55: Utilize the subtask production conflict data to perform flexible scheduling processing on the priority production subtask network data to generate conflict-free production subtask network data.

7. The data processing method based on the industrial Internet data platform according to claim 6 is characterized in that: Step S54 includes the following steps: Step S541: Perform production time series analysis on the priority production subtask network data to obtain production subtask time series data; Step S542: Perform time constraint processing according to the production subtask timing data, and perform subtask time conflict detection to generate subtask time conflict data; Step S543: Perform production resource demand analysis on the priority production subtask network data to generate production resource demand data; Step S544: matching the production resource demand data with the production resource capacity by optimizing the production capacity matrix data to generate resource capacity matching data; Step S545: Perform production resource conflict detection according to the resource capability matching data to obtain subtask production resource conflict data; Step S546: Perform production process constraint violation detection on the priority production subtask network data through the production process constraint knowledge graph to obtain subtask production constraint violation data.

8. The data processing method based on the industrial Internet data platform according to claim 6 is characterized in that: Step S55 includes the following steps: Step S551: performing time constraint relaxation processing on the subtask time conflict data to generate subtask production relaxation time data; Step S552: rescheduling the priority production subtask network data according to the subtask production slack time data, and screening the optimal time conflict scheduling according to the preset time flexibility priority adjustment rule to generate the optimal time conflict scheduling data; Step S553: ​​performing alternative production resource analysis on the subtask production resource conflict data to generate alternative production resource data; Step S554: rescheduling the network data of the priority production subtasks to resolve resource conflicts through the replaceable production resource data, and adjusting the load balancing optimization to generate optimal resource conflict allocation data; Step S555: performing constraint relaxation determination on the subtask production constraint violation data, and generating loosenable constraint data; Step S556: adjusting the constraint parameters of the priority production subtask network data through the releasable constraint data, and performing production constraint risk assessment to generate low-risk constraint adjustment data; Step S557: Use low-risk constraint adjustment data, optimal resource conflict allocation data, and optimal time conflict scheduling data to eliminate conflicts in the priority production subtask network data, perform global conflict verification, and generate conflict-free production subtask network data.

Images