A method and system for collaborative management of intelligent interconnected factory resources

Abstract

The present invention discloses a method and system for collaborative management of resources in an intelligent interconnected factory, which relates to the technical field of collaborative management of factory resources. The method obtains historical production monitoring data of each resource production line, determines the production efficiency of semi-finished products and constructs resource particles, generates a historical resource particle fusion monitoring window, and analyzes the historical fusion performance of resource particles; establishes a correlation between historical production monitoring data, fusion performance, abnormal impact time and volume, and forms a historical abnormality judgment data group; constructs a real-time fusion monitoring window based on the real-time semi-finished product production efficiency, analyzes the real-time fusion performance, compares it with the historical fusion performance, and determines the adapted historical production monitoring data; uses the abnormal impact time and volume of the adapted historical data as early warning information to issue an early warning to the production line. The present invention improves the resource collaboration efficiency and production stability of the intelligent interconnected factory and reduces the risk of production interruption by dynamically monitoring resource allocation and accurately identifying abnormal patterns.

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative management of factory resources, and in particular to a method and system for collaborative management of resources in an intelligent interconnected factory. Background Art

[0002] With the rapid development of Industry 4.0 and smart manufacturing, smart, connected factories have become a vital component of modern manufacturing. Leveraging the Internet of Things, big data analytics, and artificial intelligence (AI), these factories enable the coordinated operation of multiple production lines and efficient resource allocation. However, in actual production, factories face challenges such as difficulty coordinating resources, low production efficiency, and delayed response to unexpected events due to fluctuations in the efficiency of semi-finished products, complex resource dependencies, and unpredictable abnormal events.

[0003] Traditional resource management methods typically rely on static scheduling or experience-based dynamic adjustments, making them ill-suited to the real-time and complex nature of production data in smart, connected factories. For example, existing technologies often monitor production lines based on a single metric (such as output or equipment operating status), lacking comprehensive analysis of resource allocation ratios, work-in-progress volume requirements, and the impact of anomalies. Furthermore, anomaly detection often relies on simple threshold judgments or manual intervention, making it difficult to accurately predict the impact time and volume of anomalies. This can lead to delayed warnings or misjudgments, further impacting production continuity and resource utilization efficiency. Summary of the Invention

[0004] The purpose of the present invention is to provide a resource collaborative management method and system capable of collaboratively monitoring semi-finished products on a production line.

[0005] The present invention discloses a method for collaborative management of resources in an intelligent interconnected factory, comprising:

[0006] Step S100: Obtain historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data, and determine the historical fusion performance of all resource particles within the historical resource particle fusion monitoring window;

[0007] Step S200: Determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish a correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain a historical abnormal impact judgment data group;

[0008] Step S300: Obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles therein. Compare and analyze the real-time fusion performance using different historical fusion performances to determine several matching historical fusion performances. Compare the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine the matching historical production monitoring data.

[0009] In step S400, the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data are identified as resource collaborative warning information, and a warning is issued to the corresponding resource production line.

[0010] In an embodiment of the present invention, a method for constructing a resource particle fusion monitoring window includes:

[0011] Step S101: Identify the semi-finished products corresponding to different resource production lines, determine the volume required for each semi-finished product to be the final product, and calculate the required volume multiple of each type of semi-finished product relative to the finished product.

[0012] Step S102: Determine the resource particles generated per unit time based on the semi-finished product production efficiency and the required volume multiple of the resource production line;

[0013] Step S103: Construct a time monitoring window for resource particles and construct a resource particle ratio sequence in the order of consecutive time nodes. Each resource particle ratio sequence includes different resource particles corresponding to the time node, and the number of resource particles is configured according to the number generated at the corresponding time node.

[0014] Step S104: a monitoring time length is preset for the time monitoring window to obtain a resource particle fusion monitoring window.

[0015] In an embodiment of the present invention, a method for determining the fusion performance of resource particles within a resource particle fusion monitoring window includes:

[0016] Step S105: For the resource particle ratio of each type of resource particle in each resource particle ratio sequence to all resource particles, determine the fluctuation characteristics of the resource particle ratio between different resource particle ratio sequences, and determine the total number of monitored resource particles for all resource particles within a preset time period;

[0017] Step S106 : identifying the fluctuation characteristics of the resource particle ratio change and the total number of monitored resource particles as fusion performance.

[0018] In an embodiment of the present invention, a method for comparing and analyzing historical fusion performance and real-time fusion performance includes:

[0019] Step S301: Determine the historical resource particle ratio change fluctuation characteristics and the total number of historically monitored resource particles as shown in the historical fusion performance, determine the real-time resource particle ratio change fluctuation characteristics and the total number of real-time monitored resource particles as shown in the real-time fusion performance, and compare the fluctuation difference characteristics between the resource particle ratio change fluctuation characteristics and the total number of monitored resource particles.

[0020] The fluctuation difference characteristics include the fluctuation period of the proportion change of each type of resource particles and the average proportion change fluctuation amplitude;

[0021] In step S302, if the difference in the proportion change fluctuation period and the average proportion change fluctuation value between each type of resource particles is within a preset range, and the total number difference between the total number of monitored resource particles is also within a preset range, then it is determined that the historical fusion performance is consistent with the real-time fusion performance.

[0022] In an embodiment of the present invention, a method for comparing historical production monitoring data with real-time production monitoring data includes:

[0023] Step S303: historical production monitoring data marked with abnormal production status are identified, and the task volume requirements corresponding to each historical production monitoring data are determined. Based on the equality of the task volume requirements, the historical production monitoring data are summarized and classified to form a number of comparison production monitoring data groups. The historical production monitoring data marked with abnormal production status in the comparison production monitoring data groups are compared with other historical production monitoring data, and based on the comparison results, a number of focused production monitoring data are determined.

[0024] In step S304, the task requirements of the real-time production monitoring data are determined, and several corresponding comparison production monitoring data groups are determined, and the respective focused production monitoring data are determined. The focused production monitoring data and the corresponding data in the real-time production monitoring data are compared to determine the degree of compatibility between the two.

[0025] In an embodiment of the present invention, a method for determining the production monitoring data to be focused on includes:

[0026] Step S3031, parameterizing each type of data in the historical production monitoring data, and constructing monitoring data parameter curves respectively;

[0027] Step S3032: aligning the data parameter curves of the historical production monitoring data marked with abnormal production status and those not marked with abnormal production status in time, and segmenting them according to equivalent production capacity cycles, thereby obtaining several types of normal monitoring data parameter curve segment groups and several types of abnormal monitoring data parameter curve segment groups, wherein each normal monitoring data parameter curve segment group includes several normal monitoring data parameter curve segments, and each abnormal monitoring data parameter curve segment group includes several abnormal monitoring data parameter curve segments;

[0028] Step S3033: Merge the normal monitoring data parameter curve segment groups and the abnormal monitoring data parameter curve segment groups of the same data type that belong to the same comparison production monitoring data group to obtain a normal monitoring data parameter curve segment set and an abnormal monitoring data parameter curve segment set;

[0029] Step S3034: Performing curvature change analysis and curve parameter value analysis on the monitoring parameter curve segments in the monitoring data parameter curve segment set. The curvature change analysis includes determining the proportion of the first duration of the curvature change in different curvature intervals and constructing a corresponding relationship sequence from the curvature interval to the first duration. The curve parameter value analysis includes determining the proportion of the second duration of the curve parameter values ​​in different curve parameter value intervals and constructing a corresponding relationship sequence from the curve parameter value interval to the second duration.

[0030] Step S3035, based on the comparison relationship between the abnormal monitoring data parameter curve segment set and the normal monitoring data parameter curve segment set of the same data type, establish a comparison relationship of the curvature interval~first duration correspondence sequence, and a comparison relationship of the curve parameter value interval~second duration proportion correspondence sequence, and perform a comparison analysis based on the comparison relationship. Based on the analysis results, determine the prominence of the abnormal monitoring data parameter curve segment set relative to the normal monitoring data parameter curve segment set. If the prominence is greater than or equal to the preset value, it is determined that the monitoring data corresponding to the data type is production-focused monitoring data.

[0031] In the embodiment of the present invention, comparing the historical production monitoring data and the real-time production monitoring data corresponding to each determined historical fusion performance includes:

[0032] Step S301: Determine the focused production monitoring data in the historical production monitoring data and the real-time production monitoring data, parameterize each focused production monitoring data, and construct a parameter curve for each focused monitoring data. Based on the correspondence between the historical production monitoring data and the real-time production monitoring data, compare the corresponding parameter curves of the focused monitoring data. The comparison method includes:

[0033] Step S3011: Dynamically shift the focus monitoring data parameter curves relative to each other, and extract in real time the curve mapping area change between the two during the shift process. The curve mapping area change relative to the unit mapping area increment at different time nodes is confirmed, and a unit mapping area increment curve is constructed. If, within a preset time period, the value corresponding to the unit mapping area increment curve is less than or equal to a preset value, and the curvature accumulation value is less than or equal to a preset value, the dynamic shift between the focus monitoring data parameter curves is terminated, and the focus monitoring data parameter curves are deemed to be consistent with each other.

[0034] Step S3012: If all the relative parameter curves of the focused monitoring data match each other, it is determined that the historical production monitoring data and the real-time production monitoring data are compatible.

[0035] In an embodiment of the present invention, there is also disclosed an intelligent interconnected factory resource collaborative management system, comprising:

[0036] The first module is used to obtain the historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, and construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data. The historical fusion performance of all resource particles within the historical resource particle fusion monitoring window is determined;

[0037] The second module is used to determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish the correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain the historical abnormal impact judgment data group;

[0038] The third module is used to obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles within it. It then compares and analyzes the real-time fusion performance using different historical fusion performances to determine several matching historical fusion performances. It then compares the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine the matching historical production monitoring data.

[0039] The fourth module is used to identify the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data as resource collaborative warning information to issue an early warning to the corresponding resource production line.

[0040] The present invention discloses a method and system for collaborative management of resources in an intelligent interconnected factory, which relates to the technical field of collaborative management of factory resources. The method obtains historical production monitoring data of each resource production line, determines the production efficiency of semi-finished products and constructs resource particles, generates a historical resource particle fusion monitoring window, and analyzes the historical fusion performance of resource particles; establishes a correlation between historical production monitoring data, fusion performance, abnormal impact time and volume, and forms a historical abnormality judgment data group; constructs a real-time fusion monitoring window based on the real-time semi-finished product production efficiency, analyzes the real-time fusion performance, compares it with the historical fusion performance, and determines the adapted historical production monitoring data; uses the abnormal impact time and volume of the adapted historical data as early warning information to issue an early warning to the production line. The present invention improves the resource collaboration efficiency and production stability of the intelligent interconnected factory and reduces the risk of production interruption by dynamically monitoring resource allocation and accurately identifying abnormal patterns.

[0041] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a method step diagram of a method for collaborative management of intelligent interconnected factory resources disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0044] The following will be combined with the accompanying drawings and specific embodiments to clearly and completely describe the technical solutions of the present invention. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and cannot be understood as limiting the scope of protection of the present invention. Those skilled in the art in this field can make some non-essential improvements and adjustments based on the content of the present invention described below. In the present invention, unless otherwise clearly specified and limited, the technical terms used in the present invention should have the common meanings understood by those skilled in the art of the present invention.

[0045] Example:

[0046] The present invention discloses a method for collaborative management of resources in an intelligent interconnected factory. Figure 1 ,include:

[0047] Step S100: Obtain the historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, and construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data, and determine the historical fusion performance of all resource particles within the historical resource particle fusion monitoring window.

[0048] Step S100 aims to construct a baseline model for resource allocation and production efficiency based on historical production data, providing a foundation for subsequent anomaly detection and real-time monitoring. The system first collects production monitoring data from each resource production line during its past operation, including indicators such as output, equipment operating status, raw material consumption, and production cycle time, comprehensively reflecting the production line's operational status. By analyzing this data, the system quantifies the production efficiency of each production line in semi-finished products, defined as the number of qualified semi-finished products produced per unit time (for example, 100 semi-finished products per hour). To facilitate analysis, the system introduces the concept of "resource particles." Each resource particle represents the output of semi-finished products per unit time, specifically the number of semi-finished products produced within a certain time period (for example, one hour), abstracted as a quantifiable virtual data unit. For example, if a production line produces 100 semi-finished products per hour, 100 resource particles are generated, each representing the output of one semi-finished product. Based on the semi-finished product production efficiency of each production line, the system generates a corresponding number of resource particles and constructs a "historical resource particle fusion monitoring window." This window is a time-dimensional analysis framework that integrates the resource particles of all production lines within the same time period to form a sequence of resource particle ratios at consecutive time nodes. The number and ratio of resource particles across different production lines are recorded sequentially, reflecting the state of resource allocation. The system analyzes fluctuation characteristics (e.g., period and amplitude) and total changes in the ratio of resource particles within the window to determine "historical fusion performance." This performance quantifies the stability and efficiency of resource coordination during historical production. For example, a stable particle ratio indicates balanced resource allocation, while sharp fluctuations may indicate historical anomalies (such as equipment failure). This step simplifies complex production dynamics into an actionable data model by representing the output of resource particles, facilitating subsequent comparison and anomaly identification. The historical fusion performance serves as a benchmark for real-time monitoring, ensuring the system can detect deviations in production status. This process requires processing large-scale time series data and relies on efficient data analysis techniques, laying the foundation for dynamic resource management in smart, connected factories.

[0049] Step S200: determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish the correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain the historical abnormal impact judgment data group.

[0050] Step S200 focuses on extracting anomaly features from historical production monitoring data and building a correlation model to provide a basis for real-time anomaly prediction. The system first analyzes historical production monitoring data to identify abnormal events, such as equipment failures, sudden drops in production, or resource shortages, by detecting significant deviations in the data (e.g., lower-than-expected production or abnormal equipment parameters). For each anomaly, the system quantifies its "anomaly impact time," the duration from the occurrence of the anomaly to its recovery (e.g., a two-hour downtime), and its "anomaly impact volume," the specific impact on production, such as the number of semi-finished products reduced in production (expressed in resource particles; for example, a reduction of 200 resource particles equates to 200 semi-finished products). These attributes are extracted from production logs through statistical analysis and time series decomposition and cross-validated with equipment status and resource consumption data. The system then correlates the anomaly features with the historical fusion representation generated in step S100 (resource particle ratio fluctuations and total volume characteristics) and historical production monitoring data. For example, an anomaly might be associated with a sudden drop in the resource particle ratio for a particular production line, indicating an imbalance in resource allocation.

[0051] Step S300: Obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles therein. Use different historical fusion performances to compare and analyze the real-time fusion performance to determine several matching historical fusion performances. Compare the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine the matching historical production monitoring data.

[0052] Step S300 matches real-time production status monitoring with historical anomaly patterns, with the core goal of rapidly identifying potential anomalies. The system first collects current semi-finished product production efficiency data for each production line, covering real-time output and resource utilization, reflecting the current operating status. Based on the resource particle framework established in step S100, the system converts real-time efficiency data into resource particles, with each particle representing the output of semi-finished products per unit time (for example, 150 semi-finished products produced per hour corresponds to 150 resource particles). The system constructs a "real-time resource particle fusion monitoring window," integrating the resource particles from each production line to form a sequence of resource particle ratios at consecutive time points, recording the number and ratio of particles. This sequence is analyzed to generate a "real-time fusion performance," which includes the fluctuation characteristics (period, amplitude) and total volume changes of the resource particle ratios. The system compares the real-time fusion performance with the historical fusion performance from step S100 and uses feature matching algorithms (such as time series similarity or statistical distance) to identify similar historical fusion performances based on characteristics such as fluctuation period, amplitude, and total volume differences. For each matched historical fusion performance, the system extracts the corresponding historical production monitoring data and performs a detailed comparison with real-time production monitoring data. This comparison includes task volume requirements, production curves, and equipment parameters, identifying the historical production monitoring data that best matches the current scenario. The resource particle yield definition makes real-time efficiency quantifiable and intuitive, facilitating fluctuation analysis. Multi-level comparison (fusion performance first, production data second) improves matching accuracy and reduces misjudgments. The matched historical data represents the scenario closest to the current state and typically contains known anomalies, providing a basis for early warning. This step relies on real-time data processing and feature extraction to rapidly respond to dynamic environments. Its innovation lies in connecting historical and real-time data through fusion performance, providing a systematic solution for anomaly prediction in complex scenarios and ensuring the stability of the interconnected factory.

[0053] In step S400, the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data are identified as resource collaborative warning information, and a warning is issued to the corresponding resource production line.

[0054] Step S400 converts historical anomaly characteristics into real-time warning information to prevent production disruptions. Based on the adapted historical production monitoring data from step S300, the system extracts the anomaly impact duration (e.g., a three-hour downtime) and impact volume (e.g., a reduction of 500 resource particles, equivalent to 500 semi-finished products) from the data set from step S200. These attributes are encapsulated as "resource collaborative warning information," clarifying the type, duration, and impact range of the potential anomaly. The system transmits this information to relevant production lines through alarm signals, interface prompts, or control system integration, prompting them to take action, such as adjusting resources or maintaining equipment. The resource particle yield definition allows for intuitive representation of volume in terms of semi-finished product numbers, making it easier for operators to understand. The accuracy of warnings stems from the prediction of historical anomaly patterns to ensure relevance to the current scenario. Implementation requires integration with the factory communication system to ensure timely information transmission and prioritize alerts by anomaly severity. This step converts analysis results into instructions, closing the data-action loop, reducing disruption risks and improving resource efficiency. Its success relies on the accuracy of historical data and real-time processing capabilities, demonstrating the system's practicality in dynamic environments. The early warning mechanism uses history to guide current operations, improves response timeliness, enhances the adaptability of interconnected factories to complex scenarios, and provides support for efficient production.

[0055] In an embodiment of the present invention, a method for constructing a resource particle fusion monitoring window includes:

[0056] Step S101: Identify the semi-finished products corresponding to different resource production lines, determine the volume required for each semi-finished product to be the final product, and calculate the required volume multiple of each type of semi-finished product relative to the finished product.

[0057] Step S102: Determine the resource particles generated per unit time based on the semi-finished product production efficiency and the required volume multiple of the resource production line;

[0058] Step S103: Construct a time monitoring window for resource particles and construct a resource particle ratio sequence in the order of consecutive time nodes. Each resource particle ratio sequence includes different resource particles corresponding to the time node, and the number of resource particles is configured according to the number generated at the corresponding time node.

[0059] Step S104: a monitoring time length is preset for the time monitoring window to obtain a resource particle fusion monitoring window.

[0060] In an embodiment of the present invention, a method for determining the fusion performance of resource particles within a resource particle fusion monitoring window includes:

[0061] Step S105: For the resource particle ratio of each type of resource particle in each resource particle ratio sequence to all resource particles, determine the fluctuation characteristics of the resource particle ratio between different resource particle ratio sequences, and determine the total number of monitored resource particles for all resource particles within a preset time period;

[0062] Step S106 : identifying the fluctuation characteristics of the resource particle ratio change and the total number of monitored resource particles as fusion performance.

[0063] In an embodiment of the present invention, a method for comparing and analyzing historical fusion performance and real-time fusion performance includes:

[0064] Step S301: Determine the historical resource particle ratio change fluctuation characteristics and the total number of historically monitored resource particles as shown in the historical fusion performance, determine the real-time resource particle ratio change fluctuation characteristics and the total number of real-time monitored resource particles as shown in the real-time fusion performance, and compare the fluctuation difference characteristics between the resource particle ratio change fluctuation characteristics and the total number of monitored resource particles.

[0065] The fluctuation difference characteristics include the fluctuation period of the proportion change of each type of resource particles and the average proportion change fluctuation amplitude;

[0066] In step S302, if the difference in the proportion change fluctuation period and the average proportion change fluctuation value between each type of resource particles is within a preset range, and the total number difference between the total number of monitored resource particles is also within a preset range, then it is determined that the historical fusion performance is consistent with the real-time fusion performance.

[0067] In an embodiment of the present invention, a method for comparing historical production monitoring data with real-time production monitoring data includes:

[0068] Step S303, determine the historical production monitoring data marked with abnormal production status, determine the task volume requirement corresponding to each historical production monitoring data, summarize and classify the historical production monitoring data based on the equality of the task volume requirement, form several comparison production monitoring data groups, compare the historical production monitoring data marked with abnormal production status with others in the comparison production monitoring data groups, and determine several focused production monitoring data based on the comparison results.

[0069] Step S303 aims to systematically analyze historical production monitoring data to identify the most representative data for anomaly detection, providing an accurate reference for subsequent comparison with real-time data. The system first identifies records from the historical production monitoring data that are marked with abnormal production conditions. These anomalies may include equipment failure, sudden drop in production, or resource shortages, and distinguishes them using anomaly labels (such as fault codes or production anomaly flags). Each record of historical production monitoring data contains indicators such as production output, equipment status, and resource consumption. The system then extracts the "task demand," which is the expected semi-finished product output for the production task (represented by resource particles, with each particle corresponding to the semi-finished product output per unit time; for example, 100 semi-finished products per hour corresponds to 100 resource particles). The task demand reflects the production goals and resource allocation requirements of the production line and is an important basis for comparing historical and real-time data. Based on the equivalence of task demand, the system categorizes the historical production monitoring data, grouping data with similar task demand into the same "comparison production monitoring data group." For example, data with a task demand of 500 semi-finished products per hour are grouped together to ensure comparability of data within the group with respect to production targets. Within each comparison production monitoring data set, the system conducts a detailed comparison of historical production monitoring data marked as abnormal with other non-anomalous data, analyzing their distinguishing characteristics, such as the shape of the yield curve, fluctuations in resource particle ratio (based on output per unit time), or changes in equipment operating parameters. Through this comparison, the system identifies unique patterns in the abnormal data relative to normal data. For example, abnormal data might show a sharp fluctuation in resource particle ratio or a sudden drop in output. Based on these differences, the system identifies "focused production monitoring data," a subset of data with prominent characteristics for anomaly detection.

[0070] In step S304, the task requirements of the real-time production monitoring data are determined, and several corresponding comparison production monitoring data groups are determined, and the respective focused production monitoring data are determined. The focused production monitoring data and the corresponding data in the real-time production monitoring data are compared to determine the degree of compatibility between the two.

[0071] Step S304 accurately matches real-time production monitoring data with historical data to assess the similarity between the current production status and historical anomaly patterns, providing a basis for early warning. The system first collects real-time production monitoring data, including current output, equipment status, and resource consumption. Based on this data, it determines the real-time "task demand," or the expected semi-finished product output for the current production task (expressed in resource particles; for example, 200 semi-finished products per hour corresponds to 200 resource particles). Using the task demand as a key metric for matching, the system selects corresponding historical data sets from the comparison production monitoring data sets generated in step S303 based on the real-time task demand, ensuring comparability between historical and real-time data on production targets. For example, if the real-time task demand is 500 semi-finished products per hour, the system selects comparison production monitoring data sets with similar historical task demand. Within this selected data set, the system further extracts the "focused production monitoring data" determined in step S303, which represent typical characteristics of anomaly conditions.

[0072] In an embodiment of the present invention, a method for determining the production monitoring data to be focused on includes:

[0073] Step S3031: Parameterize each type of data in the historical production monitoring data and construct monitoring data parameter curves respectively.

[0074] In step S3031, the system parameterizes sensor node data from historical production monitoring data, constructing monitoring data parameter curves and converting the raw data from different sensors into a quantifiable and analyzable format. Sensor data includes indicators such as temperature (e.g., equipment bearing temperature), pressure (e.g., hydraulic system pressure), vibration (e.g., motor vibration frequency), or current (e.g., equipment load current), reflecting the real-time status of equipment operation and potential anomalies. The parameterization process converts this heterogeneous data into standardized numerical representations. For example, temperature is normalized to a deviation from a baseline temperature, or vibration frequency is quantified to a statistical characteristic value of the amplitude (e.g., root mean square value). To ensure comparability between different sensor data, normalization or standardization is typically employed. Subsequently, the system plots a monitoring data parameter curve, with time as the horizontal axis and the parameterized value as the vertical axis. For example, a curve depicting the change in vibration intensity over time for a particular sensor node is shown. The principle behind this step is to convert the dynamic behavior of sensor data into a continuous time series representation, facilitating analysis of trends, periodic fluctuations, or anomalies (e.g., sudden increases in vibration). By generating smooth parametric curves, the system supports mathematical analysis (such as frequency analysis or differentiation), laying the foundation for distinguishing normal from abnormal sensor data patterns. This process is the core starting point for subsequent anomaly detection, ensuring that sensor data can be structured for pattern recognition.

[0075] Step S3032, the data parameter curves of the historical production monitoring data marked with abnormal production status and unmarked abnormal production status are aligned in time, and divided according to equivalent production capacity cycles, to obtain several types of normal monitoring data parameter curve segment groups and several types of abnormal monitoring data parameter curve segment groups, each normal monitoring data parameter curve segment group contains several normal monitoring data parameter curve segments, and each abnormal monitoring data parameter curve segment group contains several abnormal monitoring data parameter curve segments.

[0076] Step S3032 time-aligns the sensor data parameter curves marked as abnormal (e.g., equipment failure or abnormal vibration indicated by sensor data) and those not marked as abnormal. These curves are segmented based on equivalent production cycles to generate comparable normal and abnormal curve segments. Time alignment ensures that curves from different sensor nodes or production batches are synchronized to a unified time frame, eliminating the impact of data acquisition time differences. For example, this can be achieved by aligning production task start times (e.g., 08:00 AM per shift). After alignment, the system segments the curves into multiple segments based on production cycles (e.g., an 8-hour shift or equipment operation cycle). Each segment represents an independent production run unit, ensuring that comparisons are based on similar equipment operating conditions. After segmentation, the system categorizes the curve segments into two groups: normal monitoring data parameter curve segment groups (from sensor data without abnormalities, such as stable temperature or vibration) and abnormal monitoring data parameter curve segment groups (from abnormal data, such as sudden temperature rise or abnormal vibration). Each group contains multiple segments of the same type of sensor data (e.g., vibration or current). This step is based on isolating sensor data under comparable conditions through time alignment and period segmentation, reducing noise introduced by differences in production schedules or acquisition times. This structured processing facilitates subsequent analysis of the differences between normal and abnormal sensor data patterns, providing a reliable data foundation for accurate anomaly detection.

[0077] Step S3033: Merge the normal monitoring data parameter curve segment groups and abnormal monitoring data parameter curve segment groups that belong to the same comparison production monitoring data group and have the same data type to obtain a normal monitoring data parameter curve segment set and an abnormal monitoring data parameter curve segment set.

[0078] In step S3033, the system merges normal and abnormal sensor data parameter curve segment groups of the same type that belong to the same comparison production monitoring data group to form normal and abnormal monitoring data parameter curve segment sets. A comparison production monitoring data group consists of sensor data with similar workload requirements (for example, sensor data from equipment that supports the production of 500 semi-finished products per hour), ensuring that the curve segments within the group are comparable in terms of production targets and equipment load. Within each group, curve segments of the same type of sensor data (such as vibration frequency or temperature) are aggregated to represent the normal and abnormal states of the same comparison group. For example, all normal vibration frequency curve segments are merged into a normal monitoring data parameter curve segment set, while abnormal vibration frequency curve segments form the corresponding abnormal set. The merging process groups these segments, preserving their temporal and parameter characteristics, to form a comprehensive dataset covering multiple production runs. The principle behind this step is to enhance the statistical robustness of the data through aggregation, enabling the system to analyze the collective behavior of sensor data across multiple runs, rather than a single instance. The merged curve segment set provides a larger sample size, making it easier to identify significant patterns between normal and abnormal conditions (such as the spike characteristics of the vibration curve under abnormal conditions), laying the foundation for subsequent comparative analysis and ensuring the reliability and representativeness of the results.

[0079] Step S3034, performing curvature change analysis and curve parameter value analysis on the monitoring parameter curve segments in the monitoring data parameter curve segment set, the curvature change analysis includes determining the first duration ratio of the curvature change in different curvature intervals, and constructing a curvature interval~first duration ratio correspondence sequence, the curve parameter value analysis includes determining the second duration ratio of the curve parameter values ​​in different curve parameter value intervals, and constructing a curve parameter value interval~second duration ratio correspondence sequence.

[0080] Step S3034 performs curvature change analysis and curve parameter value analysis on the normal and abnormal monitoring data parameter curve segments to quantify the temporal dynamics and amplitude characteristics of the sensor data. Curvature change analysis assesses the rate of change of the curve shape, reflecting the temporal fluctuation characteristics of sensor parameters (such as vibration frequency or temperature). The system calculates the curvature of each curve segment and categorizes it into predefined curvature intervals (e.g., low, medium, and high curvature). It also calculates the first duration percentage (i.e., the percentage of time each curvature interval spends within the segment set), generating a sequence of corresponding curvature interval second duration percentages. This step decomposes sensor data behavior into two dimensions: dynamic (curvature, capturing parameter abrupt changes or recovery) and static (parameter value, reflecting parameter levels). These sequences represent sensor operating patterns in a structured manner, facilitating the identification of abnormal characteristics, such as vibration curvature spikes or prolonged high-temperature conditions caused by faults. This quantitative analysis provides key metrics for subsequent comparisons between normal and abnormal conditions, supporting accurate abnormal pattern detection.

[0081] Step S3035, based on the comparison relationship between the abnormal monitoring data parameter curve segment set and the normal monitoring data parameter curve segment set of the same data type, establish a comparison relationship of the curvature interval~first duration correspondence sequence, and a comparison relationship of the curve parameter value interval~second duration proportion correspondence sequence, and perform a comparison analysis based on the comparison relationship. Based on the analysis results, determine the prominence of the abnormal monitoring data parameter curve segment set relative to the normal monitoring data parameter curve segment set. If the prominence is greater than or equal to the preset value, it is determined that the monitoring data corresponding to the data type is production-focused monitoring data.

[0082] The expression for calculating the prominence is: ;

[0083] in, For prominence, Highlight weight coefficients for preset curvature changes, Highlight weight coefficients for preset curve parameter values, is the first duration difference highlight judgment function, if the curvature interval ~ first duration correspondence sequence, the If the first duration difference corresponding to the curvature interval is greater than or equal to the preset value, then Output 1, otherwise output 0. is the number of curvature intervals in the curvature interval~first duration correspondence sequence, is the first duration difference performance impact adjustment coefficient, is the first duration difference performance effect adjustment constant, The second duration difference highlight judgment function, if the curve parameter value interval ~ the second duration ratio correspondence sequence, the If the difference in the second duration of the curve parameter value interval is greater than or equal to the preset value, then Output 1, otherwise output 0. is the adjustment coefficient for the second duration difference performance impact, is the adjustment constant for the impact of the second duration difference performance, and N is the number of curve parameter value intervals in the curve parameter value interval~second duration proportion corresponding relationship sequence.

[0084] In the embodiment of the present invention, comparing the historical production monitoring data and the real-time production monitoring data corresponding to each determined historical fusion performance includes:

[0085] Step S301: Determine the focused production monitoring data in the historical production monitoring data and the real-time production monitoring data, parameterize each focused production monitoring data, and construct a parameter curve for each focused monitoring data. Based on the correspondence between the historical production monitoring data and the real-time production monitoring data, compare the corresponding parameter curves of the focused monitoring data. The comparison method includes:

[0086] Step S3011: Dynamically shift the focus monitoring data parameter curves relative to each other, and extract in real time the curve mapping area change between the two during the shift process. The curve mapping area change relative to the unit mapping area increment at different time nodes is confirmed, and a unit mapping area increment curve is constructed. If, within a preset time period, the value corresponding to the unit mapping area increment curve is less than or equal to a preset value, and the curvature accumulation value is less than or equal to a preset value, the dynamic shift between the focus monitoring data parameter curves is terminated, and the focus monitoring data parameter curves are deemed to be consistent with each other.

[0087] In step S3012, if all the relative parameter curves of the focused monitoring data match each other, it is determined that the historical production monitoring data and the real-time production monitoring data are compatible.

[0088] In an embodiment of the present invention, there is also disclosed an intelligent interconnected factory resource collaborative management system, comprising:

[0089] The first module is used to obtain the historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, and construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data. The historical fusion performance of all resource particles within the historical resource particle fusion monitoring window is determined;

[0090] The second module is used to determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish the correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain the historical abnormal impact judgment data group;

[0091] The third module is used to obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles within it. It then compares and analyzes the real-time fusion performance using different historical fusion performances to determine several matching historical fusion performances. It then compares the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine the matching historical production monitoring data.

[0092] The fourth module is used to identify the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data as resource collaborative warning information to issue an early warning to the corresponding resource production line.

[0093] The present invention discloses a method and system for collaborative management of resources in an intelligent interconnected factory, which relates to the technical field of collaborative management of factory resources. The method obtains historical production monitoring data of each resource production line, determines the production efficiency of semi-finished products and constructs resource particles, generates a historical resource particle fusion monitoring window, and analyzes the historical fusion performance of resource particles; establishes a correlation between historical production monitoring data, fusion performance, abnormal impact time and volume, and forms a historical abnormality judgment data group; constructs a real-time fusion monitoring window based on the real-time semi-finished product production efficiency, analyzes the real-time fusion performance, compares it with the historical fusion performance, and determines the adapted historical production monitoring data; uses the abnormal impact time and volume of the adapted historical data as early warning information to issue an early warning to the production line. The present invention improves the resource collaboration efficiency and production stability of the intelligent interconnected factory and reduces the risk of production interruption by dynamically monitoring resource allocation and accurately identifying abnormal patterns.

[0094] Through the above description of the embodiments, those skilled in the art will clearly understand that the present invention can be implemented via hardware or via software combined with a necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product. This software product can be stored on a non-volatile storage medium (such as a CD-ROM, USB flash drive, or external hard drive) and includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute the methods described in various implementation scenarios of the present invention.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for collaborative management of resources in an intelligent interconnected factory, characterized in that: include: Step S100: Obtain historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data, and determine the historical fusion performance of all resource particles within the historical resource particle fusion monitoring window; Step S200: Determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish a correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain a historical abnormal impact judgment data group; Step S300: Obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles therein. Compare and analyze the real-time fusion performance using different historical fusion performances to determine several matching historical fusion performances. Compare the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine matching historical production monitoring data. Resource particles represent the semi-finished product output per unit time. Step S400: identifying the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data as resource collaborative warning information, and issuing a warning to the corresponding resource production line; Methods for constructing a resource particle fusion monitoring window include: Step S101: Identify the semi-finished products corresponding to different resource production lines, determine the volume required for each semi-finished product to be the final product, and calculate the required volume multiple of each type of semi-finished product relative to the finished product. Step S102: Determine the resource particles generated per unit time based on the semi-finished product production efficiency and the required volume multiple of the resource production line; Step S103: Construct a time monitoring window for resource particles and construct a resource particle ratio sequence in the order of consecutive time nodes. Each resource particle ratio sequence includes different resource particles corresponding to the time node, and the number of resource particles is configured according to the number generated at the corresponding time node. Step S104: a monitoring time length is preset for the time monitoring window to obtain a resource particle fusion monitoring window; The method for determining the fusion performance of resource particles within the resource particle fusion monitoring window includes: Step S105: For the resource particle ratio of each type of resource particle in each resource particle ratio sequence to all resource particles, determine the fluctuation characteristics of the resource particle ratio between different resource particle ratio sequences, and determine the total number of monitored resource particles for all resource particles within a preset time period; Step S106 : identifying the fluctuation characteristics of the resource particle ratio change and the total number of monitored resource particles as fusion performance.

2. The method for collaborative management of intelligent interconnected factory resources according to claim 1, characterized in that: Methods for comparing and analyzing historical fusion performance with real-time fusion performance include: Step S301: Determine the historical resource particle ratio change fluctuation characteristics and the total number of historically monitored resource particles as shown in the historical fusion performance, determine the real-time resource particle ratio change fluctuation characteristics and the total number of real-time monitored resource particles as shown in the real-time fusion performance, and compare the fluctuation difference characteristics between the resource particle ratio change fluctuation characteristics and the total number of monitored resource particles. The fluctuation difference characteristics include the fluctuation period of the proportion change of each type of resource particles and the average proportion change fluctuation amplitude; In step S302, if the difference in the proportion change fluctuation period and the average proportion change fluctuation value between each type of resource particles is within a preset range, and the total number difference between the total number of monitored resource particles is also within a preset range, then it is determined that the historical fusion performance is consistent with the real-time fusion performance.

3. The method for collaborative management of intelligent interconnected factory resources according to claim 1, characterized in that: Methods for comparing historical production monitoring data with real-time production monitoring data include: Step S303: historical production monitoring data marked with abnormal production status are identified, and the task volume requirements corresponding to each historical production monitoring data are determined. Based on the equality of the task volume requirements, the historical production monitoring data are summarized and classified to form a number of comparison production monitoring data groups. The historical production monitoring data marked with abnormal production status in the comparison production monitoring data groups are compared with other historical production monitoring data, and based on the comparison results, a number of focused production monitoring data are determined. In step S304, the task requirements of the real-time production monitoring data are determined, and several corresponding comparison production monitoring data groups are determined, and the respective focused production monitoring data are determined. The focused production monitoring data and the corresponding data in the real-time production monitoring data are compared to determine the degree of compatibility between the two.

4. The method for collaborative management of intelligent interconnected factory resources according to claim 3, characterized in that: Methods for determining where to focus production monitoring data include: Step S3031, parameterizing each type of data in the historical production monitoring data, and constructing monitoring data parameter curves respectively; Step S3032: aligning the data parameter curves of the historical production monitoring data marked with abnormal production status and those not marked with abnormal production status in time, and segmenting them according to equivalent production capacity cycles, thereby obtaining several types of normal monitoring data parameter curve segment groups and several types of abnormal monitoring data parameter curve segment groups, wherein each normal monitoring data parameter curve segment group includes several normal monitoring data parameter curve segments, and each abnormal monitoring data parameter curve segment group includes several abnormal monitoring data parameter curve segments; Step S3033: Merge the normal monitoring data parameter curve segment groups and the abnormal monitoring data parameter curve segment groups of the same data type that belong to the same comparison production monitoring data group to obtain a normal monitoring data parameter curve segment set and an abnormal monitoring data parameter curve segment set; Step S3034: Performing curvature change analysis and curve parameter value analysis on the monitoring parameter curve segments in the monitoring data parameter curve segment set. The curvature change analysis includes determining the proportion of the first duration of the curvature change in different curvature intervals and constructing a corresponding relationship sequence from the curvature interval to the first duration. The curve parameter value analysis includes determining the proportion of the second duration of the curve parameter values ​​in different curve parameter value intervals and constructing a corresponding relationship sequence from the curve parameter value interval to the second duration. Step S3035, based on the comparison relationship between the abnormal monitoring data parameter curve segment set and the normal monitoring data parameter curve segment set of the same data type, establish a comparison relationship of the curvature interval~first duration correspondence sequence, and a comparison relationship of the curve parameter value interval~second duration proportion correspondence sequence, and perform a comparison analysis based on the comparison relationship. Based on the analysis results, determine the prominence of the abnormal monitoring data parameter curve segment set relative to the normal monitoring data parameter curve segment set. If the prominence is greater than or equal to the preset value, it is determined that the monitoring data corresponding to the data type is production-focused monitoring data.

5. The method for collaborative management of intelligent interconnected factory resources according to claim 1, characterized in that: Comparing the historical production monitoring data and real-time production monitoring data corresponding to each determined historical fusion performance includes: Step S301: Determine the focused production monitoring data in the historical production monitoring data and the real-time production monitoring data, parameterize each focused production monitoring data, and construct a parameter curve for each focused monitoring data. Based on the correspondence between the historical production monitoring data and the real-time production monitoring data, compare the corresponding parameter curves of the focused monitoring data. The comparison method includes: Step S3011: Dynamically shift the focus monitoring data parameter curves relative to each other, and extract in real time the curve mapping area change between the two during the shift process. The curve mapping area change relative to the unit mapping area increment at different time nodes is confirmed, and a unit mapping area increment curve is constructed. If, within a preset time period, the value corresponding to the unit mapping area increment curve is less than or equal to a preset value, and the curvature accumulation value is less than or equal to a preset value, the dynamic shift between the focus monitoring data parameter curves is terminated, and the focus monitoring data parameter curves are deemed to be consistent with each other. Step S3012: If all the relative parameter curves of the focused monitoring data match each other, it is determined that the historical production monitoring data and the real-time production monitoring data are compatible.

6. An intelligent interconnected factory resource collaborative management system, characterized in that: The method for collaborative management of intelligent interconnected factory resources according to any one of claims 1 to 5 comprises: The first module is used to obtain the historical production monitoring data corresponding to each resource production line, determine the semi-finished product production efficiency corresponding to each historical production monitoring data, construct several resource particles for each semi-finished product production efficiency, and construct a historical resource particle fusion monitoring window based on the semi-finished product production efficiency corresponding to each historical production monitoring data. The historical fusion performance of all resource particles within the historical resource particle fusion monitoring window is determined; The second module is used to determine the abnormal impact time and abnormal impact volume corresponding to the historical production monitoring data, establish the correlation between the historical production monitoring data, historical fusion performance, abnormal impact time and abnormal impact volume, and obtain the historical abnormal impact judgment data group; The third module is used to obtain the real-time semi-finished product production efficiency corresponding to each resource production line, construct a real-time resource particle fusion monitoring window, and determine the real-time fusion performance of all resource particles within it. It then compares and analyzes the real-time fusion performance using different historical fusion performances to determine several matching historical fusion performances. It then compares the historical production monitoring data corresponding to each determined historical fusion performance with the real-time production monitoring data to determine the matching historical production monitoring data. The fourth module is used to identify the abnormal impact time and abnormal impact volume corresponding to the adapted historical monitoring data as resource collaborative warning information to issue an early warning to the corresponding resource production line.

Images