A digital twin enterprise collaborative management platform system
By generating a multi-dimensional mapping model through digital twin technology, and combining collaborative needs identification and process matching, the execution process is dynamically adjusted, solving the problems of data dispersion and low efficiency in traditional enterprise collaborative management, and achieving efficient collaborative management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-30
- Publication Date
- 2026-05-19
AI Technical Summary
In traditional enterprise collaborative management models, data is scattered, collaborative needs are not accurately identified, process matching deviations are large, and execution feedback is delayed, resulting in low collaborative efficiency and making it difficult to meet the needs of enterprises for efficient collaborative management.
The digital twin building module generates a multi-dimensional mapping model, which is then used in conjunction with the demand identification module for feature matching analysis. The process matching module obtains the optimal process template, and the execution feedback module makes dynamic adjustments, thereby achieving data integration, accurate demand identification, and process optimization.
It has enabled the transformation of enterprise management from static planning to dynamic optimization, improved the efficiency and quality of cross-departmental collaboration, reduced operating costs, and enhanced the enterprise's responsiveness in complex environments.
Smart Images
Figure CN120782388B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital twin technology, specifically to a digital twin enterprise collaborative management platform system. Background Technology
[0002] In the current field of enterprise management, with the deepening of digital transformation, enterprises face numerous challenges such as low cross-departmental collaboration efficiency, unreasonable resource allocation, and lagging process response. Traditional enterprise collaborative management models have the following significant shortcomings:
[0003] Data from various business systems within an enterprise is often scattered. Operational data, organizational structure data, and resource allocation data lack a unified standardized processing and integration mechanism, making it difficult to establish an accurate mapping relationship between the physical enterprise and the virtual management model. This makes it impossible to form a digital twin model that fully reflects the real state of the enterprise, thus making it difficult for management to make decisions based on a holistic perspective.
[0004] When enterprises generate collaboration needs, traditional systems struggle to quickly and accurately identify the characteristics of those needs. Due to a lack of effective utilization of historical collaboration need data, they cannot accurately pinpoint the essence of the need through feature matching analysis. This results in a low degree of matching between collaboration requests and actual process templates, frequently leading to process redundancy or missing key steps, severely impacting collaboration efficiency.
[0005] In the existing collaborative process matching process, there is often a reliance on a fixed template library. Without combining historical process matching data for dynamic optimization, it is difficult to adjust the matching strategy according to the matching coefficient based on real-time needs. This results in a deviation between the obtained process template and the actual needs, increasing the error cost of collaborative execution.
[0006] In collaborative execution, traditional management models lack the ability to collect and analyze execution data in real time, making it impossible to promptly identify execution deviations and make dynamic adjustments. Delayed feedback mechanisms in the execution process lead to problems such as unreasonable resource allocation and uncontrolled task progress that cannot be resolved in a timely manner, further exacerbating efficiency losses in enterprise collaborative management.
[0007] With the development of digital twin technology, although some enterprises have attempted to apply it to management scenarios, existing solutions mostly focus on building single-dimensional models and lack systematic integration of the entire collaborative management process (from demand identification to execution feedback), making it difficult to meet enterprises' actual needs for efficient collaboration and precise management. Therefore, there is an urgent need for an enterprise collaborative management platform system that can integrate multi-dimensional data, accurately match demands and processes, and enable dynamic adjustments to the execution process. Summary of the Invention
[0008] The purpose of this invention is to provide a digital twin enterprise collaborative management platform system to solve the problems mentioned in the background art.
[0009] To achieve the above objectives, the present invention provides the following technical solution: a digital twin enterprise collaborative management platform system, the system comprising:
[0010] The digital twin building module is used to collect basic enterprise data, establish a multi-dimensional mapping relationship between physical and virtual objects, and generate an enterprise digital twin model;
[0011] The collaborative demand identification module is used to obtain enterprise collaborative request information based on the digital twin model, extract demand feature parameters, and perform feature matching analysis in combination with historical collaborative demand data to obtain demand matching coefficients.
[0012] The process matching module is used to retrieve the enterprise collaborative process template library based on the required matching coefficient, and use the process matching feature set in the historical process matching data to perform process matching calculation on the collaborative request information to obtain the optimal collaborative process template.
[0013] The execution feedback module is used to collect collaborative execution process data based on the optimal collaborative process template, analyze the execution deviation magnitude, calculate the feedback adjustment coefficient, dynamically adjust the collaborative execution process, and form an execution feedback result.
[0014] Preferably, the process involves collecting basic enterprise data, establishing a multi-dimensional mapping relationship between physical and virtual objects, and generating a digital twin model of the enterprise, including:
[0015] Acquire enterprise operation data, organizational structure data, and resource allocation data, and process them into standardized formats respectively;
[0016] The processed operational data is mapped to the virtual operational scenario, the organizational structure data is mapped to the virtual organizational nodes, and the resource configuration data is mapped to the virtual resource units.
[0017] Integrate all related mapping relationships to generate a digital twin model of the enterprise covering operational, organizational, and resource dimensions.
[0018] Preferably, based on the digital twin model, enterprise collaboration request information is obtained, demand feature parameters are extracted, and feature matching analysis is performed in combination with historical collaboration demand data to obtain a demand matching coefficient, including:
[0019] Extract the business type, participating departments, and time requirements involved in the collaborative request from the digital twin model as demand feature parameters;
[0020] Obtain the set of historical demand characteristic parameters for similar collaborative requests within a historical time range;
[0021] Randomly select initial matching feature parameters from the set of historical demand feature parameters, and calculate the sum of the differences between the initial matching feature parameters and other historical demand feature parameters to obtain the initial matching cost;
[0022] The matching feature parameters are selected iteratively until the target matching feature parameter with the minimum matching cost is found.
[0023] Calculate the degree of matching between the current demand feature parameters and the target matching feature parameters, and use it as the demand matching coefficient.
[0024] Preferably, an initial matching feature parameter is randomly selected from the set of historical demand feature parameters, and the sum of the differences between the initial matching feature parameter and other historical demand feature parameters is calculated to obtain the initial matching cost, including:
[0025] Randomly select one historical demand feature parameter from the set of historical demand feature parameters as the initial matching feature parameter;
[0026] Calculate the differences between other historical demand feature parameters and the initial matching feature parameters in terms of business type, participating department, and time requirement, and add the differences in each dimension to obtain the total difference magnitude;
[0027] The sum of the difference magnitudes is used as the initial matching cost.
[0028] Preferably, based on the demand matching coefficient, the enterprise collaborative process template library is retrieved, and the process matching feature set in the historical process matching data is used to perform process matching calculation on the collaborative request information to obtain the optimal collaborative process template, including:
[0029] Obtain the historical process matching error set used for collaborative matching with the enterprise collaborative process template library;
[0030] Calculate the average error value of the historical process matching error set as the baseline error for the process template;
[0031] Based on the demand matching coefficient and the process template baseline error, the process matching correction coefficient is calculated.
[0032] Based on the process matching correction coefficient, the matching degree of the collaborative request information and each template in the process template library is calculated, and the template with the highest matching degree is selected as the optimal collaborative process template.
[0033] Preferably, based on the optimal collaborative process template, collaborative execution process data is collected, execution deviation amplitude is analyzed, feedback adjustment coefficients are calculated, and the collaborative execution process is dynamically adjusted to form execution feedback results, including:
[0034] Collect data on task completion progress, resource usage, and time consumption during collaborative execution as execution process data;
[0035] A machine learning method is used to train an execution feedback analyzer that includes multi-path feedback analysis, wherein the number of paths in the multi-path feedback analysis is a preset number of analyses.
[0036] Multiply the demand matching coefficient by the preset analysis quantity to obtain the actual number of analysis paths;
[0037] Within the execution feedback analyzer, a feedback analysis path of the actual number of analysis paths is randomly selected, and the execution process data is input to obtain a set of feedback adjustment coefficients.
[0038] Calculate the average value of the set of feedback adjustment coefficients, and use it as the feedback adjustment coefficient;
[0039] Based on the feedback adjustment coefficient, the task progress, resource allocation, and time arrangement during the collaborative execution process are adjusted in a positive or negative direction to form an execution feedback result.
[0040] Preferably, a machine learning method is used to train an execution feedback analyzer that includes multi-path feedback analysis, including:
[0041] Based on historical collaborative execution records, a dataset of sample execution processes was collected, and feedback adjustment coefficients under different sample execution process data were collected and labeled as a set of sample feedback adjustment coefficients.
[0042] According to the preset number of analyses, the sample execution process dataset and the sample feedback adjustment coefficient set are grouped and divided to obtain multiple sets of feedback training data;
[0043] Using the multiple sets of feedback training data respectively, a preset number of feedback analysis paths are trained based on a supervised learning algorithm, and then combined to form an execution feedback analyzer.
[0044] Preferably, enterprise operation data, organizational structure data, and resource allocation data are acquired and processed into standardized formats, including:
[0045] Real-time collection of business transaction records, production progress reports, and customer service logs from operational data via enterprise information system interfaces;
[0046] Extract departmental hierarchy, job information, and personnel affiliation from organizational structure data through the human resource management system;
[0047] The asset management system is used to obtain data on equipment distribution locations, material inventory quantities, and cash flow records from resource allocation data.
[0048] The collected operational data, organizational structure data, and resource allocation data are uniformly converted into structured tables, and duplicates and invalid fields are removed to complete the standardized format processing.
[0049] Preferably, the business type, participating departments, and time requirements involved in the collaborative request are extracted from the digital twin model as demand feature parameters, including:
[0050] Parse the collaborative request text content, identify the business keywords within it, match it with the preset business type category directory, and determine the business type;
[0051] Extract the names of the responsible and cooperating departments explicitly mentioned in the collaboration request, and confirm the participating departments by combining them with the organizational structure data in the digital twin model;
[0052] Extract the start and finish time nodes specified in the collaboration request, calculate the time span, and use it as the time requirement;
[0053] The business type, participating departments, and time span are combined to form the demand characteristic parameters.
[0054] Preferably, the historical demand feature parameter set of similar collaborative requests within a historical time range is obtained, including: filtering historical request records with the same business type as the current collaborative request, extracting the participating department information and time requirement data from each record, and organizing them into a historical demand feature parameter set containing business type, participating department, and time requirement.
[0055] Compared with the prior art, the beneficial effects of the present invention are:
[0056] In terms of data integration and model building, the digital twin construction module collects data on enterprise operations, organizational structure, and resource allocation, and performs standardized processing and multi-dimensional mapping to generate a multi-dimensional enterprise digital twin model. This process breaks down internal data silos, achieves precise mapping between physical and virtual objects, enables management to grasp the enterprise's operational status in real time and comprehensively, provides an accurate data foundation for collaborative management, and improves the scientific nature and predictability of decision-making.
[0057] In the collaborative requirement identification stage, the collaborative requirement identification module extracts feature parameters such as business type, participating departments, and time requirements from the digital twin model. It then combines these with historical collaborative requirement data for feature matching analysis to obtain a requirement matching coefficient. This mechanism, through the mining and learning of historical data, can accurately identify the essential characteristics of current requirements, avoiding the problem of requirement misunderstanding in traditional models. This improves the accuracy and efficiency of requirement identification, laying a solid foundation for subsequent process matching.
[0058] The process matching module, based on the requirement matching coefficient, retrieves the collaborative process template library and combines it with historical process matching feature sets to perform matching calculations and obtain the optimal collaborative process template. This dynamic matching mechanism fully considers the experience accumulated from historical matching data. Through the calculation of baseline error and correction coefficient, it effectively improves the fit between the process template and the current requirements, reduces process redundancy or omissions, thereby optimizing the execution efficiency of collaborative processes and reducing process management costs.
[0059] The execution feedback module collects data such as task progress, resource usage, and time consumption in real time during collaborative execution. It then calculates feedback adjustment coefficients using an execution feedback analyzer trained through machine learning, enabling dynamic adjustments to the execution process. This mechanism can promptly detect execution deviations and make positive or negative adjustments, ensuring tasks progress as planned. Simultaneously, it optimizes resource allocation and scheduling, avoiding resource waste and schedule delays, and significantly improving the accuracy and efficiency of collaborative execution.
[0060] From an overall application value perspective, this system, through the deep integration of digital twin technology and collaborative management processes, enables enterprises to shift from static planning to dynamic optimization in management. It not only improves the efficiency and quality of cross-departmental collaboration but also helps enterprises reduce operating costs, optimize resource allocation, and enhance their responsiveness and competitiveness in complex market environments through data-driven decision-making, providing strong technical support for enterprise digital transformation. Attached Figure Description
[0061] Figure 1 This is a schematic diagram illustrating the working principle of the digital twin enterprise collaborative management platform system described in this invention.
[0062] Figure 2 Sub-processes for building digital twin models;
[0063] Figure 3 The subprocess for calculating the initial matching cost;
[0064] Figure 4 Sub-processes that match the optimal collaborative process template. Detailed Implementation
[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0066] Please see Figures 1-4 The present invention relates to a digital twin enterprise collaborative management platform system, which specifically includes the following steps:
[0067] The digital twin building module collects basic enterprise data, establishes a multi-dimensional mapping relationship between physical and virtual objects, and generates an enterprise digital twin model.
[0068] The collaborative requirement identification module obtains enterprise collaborative request information based on the digital twin model, extracts requirement feature parameters, and performs feature matching analysis in combination with historical collaborative requirement data to obtain the requirement matching coefficient.
[0069] The process matching module retrieves the enterprise collaborative process template library based on the demand matching coefficient, uses the process matching feature set in the historical process matching data to perform process matching calculation on the collaborative request information, and obtains the optimal collaborative process template.
[0070] The execution feedback module collects data on the collaborative execution process based on the optimal collaborative process template, analyzes the magnitude of execution deviation, calculates the feedback adjustment coefficient, dynamically adjusts the collaborative execution process, and generates execution feedback results.
[0071] Example 1: In the specific implementation of the digital twin construction module, it is necessary to complete the collection and standardization of basic enterprise data. The system collects operational data in real time through the enterprise information system interface, specifically covering business transaction records, production progress reports, and customer service logs. Business transaction records contain detailed information for each transaction, such as transaction time, transaction partner, transaction amount, and transaction content, which accurately reflects the progress of the enterprise's business. Production progress reports record the progress of each stage of the production process, including the time nodes and completion status of raw material input, production processing, and finished product output, helping to understand the overall production rhythm. Customer service logs record detailed information such as customer inquiries, complaints, and suggestions, as well as the corresponding processing procedures and results, providing a basis for the enterprise to optimize customer service.
[0072] For organizational structure data, the system extracts departmental hierarchy, job descriptions, and personnel affiliations from the human resources management system. Departmental hierarchy clarifies the hierarchical relationships and reporting paths between various departments within the company; job descriptions define the responsibilities, authority, and requirements of each position; and personnel affiliations clearly indicate the department and position to which each employee belongs, making the company's organizational structure more transparent and standardized.
[0073] Resource allocation data acquisition relies on the asset management system, primarily including equipment location, material inventory levels, and cash flow records. Equipment location information helps companies understand the actual placement of equipment, facilitating equipment management and maintenance; material inventory levels reflect the company's real-time reserves of various materials, providing a reference for material procurement and production planning; and cash flow records detail the inflow and outflow of funds, aiding in financial analysis and cash management.
[0074] After collecting the aforementioned operational data, organizational structure data, and resource allocation data, these data need to be standardized. The system converts all collected data into a structured table format, which is characterized by its standardization, neatness, and ease of processing and analysis. During the conversion process, the system automatically removes duplicate items to avoid data redundancy and redundant storage, while also removing invalid fields to improve data quality and usability. For example, in operational data, there may be duplicate transaction records generated due to system failures or human error; these duplicates will be identified and deleted by the system. Fields irrelevant to enterprise management and decision-making, such as empty or irrelevant notes in some transaction records, will also be removed.
[0075] After standardizing the data, the next step is to establish a multi-dimensional mapping relationship between entities and virtual objects. The system associates and maps the processed operational data with virtual operational scenarios. These virtual scenarios are digital simulations of the company's actual operational processes. By mapping operational data into these virtual scenarios, real-time monitoring and simulation analysis of the company's operational processes can be achieved. For example, mapping data from production progress reports into a virtual production scenario allows for a direct view of the operational status and progress of each stage of the production line.
[0076] Organizational structure data is then mapped to virtual organizational nodes, which are digital representations of the company's actual organizational structure. Each node represents a department or position. Through this mapping, the company's organizational structure and personnel relationships can be clearly displayed in a virtual environment.
[0077] Resource allocation data is associated and mapped with virtual resource units, which are digital models of the enterprise's actual resources, including equipment, materials, and funds. Through mapping, real-time management and optimized allocation of resources can be achieved.
[0078] The system integrates all related mapping relationships to generate a digital twin model of the enterprise covering operational, organizational, and resource dimensions. This model is a digital mirror of the entire enterprise, comprehensively and accurately reflecting its actual situation. In the operational dimension, the model can simulate the enterprise's business processes, production processes, and customer service; in the organizational dimension, the model can display the enterprise's organizational structure, staffing, and job relationships; in the resource dimension, the model can manage the enterprise's equipment, materials, and funds. Through this digital twin model, enterprises can conduct various simulations and analyses in a virtual environment, providing strong support for enterprise decision-making and management. For example, enterprises can simulate new business processes in the model to assess their feasibility and effectiveness; they can simulate different organizational restructuring plans and select the optimal solution; and they can simulate resource reconfiguration to improve resource utilization efficiency.
[0079] Example 2: In the implementation of the collaborative requirement identification module, it is necessary to extract the business type, participating departments, and time requirements involved in the collaborative request from the digital twin model to form requirement feature parameters. The system parses the text content of the collaborative request and identifies business keywords through natural language processing technology. For example, keywords such as "raw material procurement" and "supplier management" are extracted from the procurement collaborative request. These keywords are then matched with a preset business type classification directory, which covers all business areas of the enterprise, such as manufacturing, marketing, human resources, and financial management, to determine the specific business type.
[0080] For extracting participating departments, the system will extract the names of the responsible and cooperating departments explicitly mentioned in the collaboration request. For example, if the request text mentions "the purchasing department will lead, with the finance and warehousing departments cooperating to complete the raw material procurement process," the system will identify the purchasing department as the responsible department and the finance and warehousing departments as the cooperating departments. Simultaneously, by combining this with the organizational structure data already constructed in the digital twin model, including departmental hierarchical relationships, job positions, and personnel affiliations, the system confirms the specific position and responsibilities of these departments within the enterprise's organizational structure, ensuring the accuracy of the participating department information.
[0081] Regarding time requirements, the system extracts the start and finish time nodes specified in the collaboration request, such as "start on June 1, 2025, and finish before June 30, 2025." By calculating the time span between these two time nodes, the specific time requirement is obtained, as shown in the example above where the time span is 30 days. Afterward, the system combines information from these three dimensions—business type, participating departments, and time span—to form complete requirement characteristic parameters.
[0082] The system needs to obtain a set of historical demand feature parameters for similar collaborative requests within a historical timeframe. Specifically, the system will filter all historical request records of the same business type from the historical database based on the business type of the current collaborative request. For example, if the current request belongs to the "raw material procurement" business type, the system will retrieve all records in the historical database marked with "raw material procurement" as the business type. For each historical record that meets the criteria, the system extracts the information of the participating departments, including the responsible department and cooperating departments, as well as the time requirement data, such as start time, completion time, and time span. This information is then organized in a unified format to form a set of historical demand feature parameters containing three dimensions: business type, participating departments, and time requirements.
[0083] After obtaining the set of historical demand feature parameters, the system needs to randomly select an initial matching feature parameter from this set and calculate the sum of the differences between this initial matching feature parameter and other historical demand feature parameters to obtain the initial matching cost. In practice, the system randomly selects one historical demand feature parameter from the set of historical demand feature parameters as the initial matching feature parameter. This selected record can be any historical record to ensure the randomness of the selection.
[0084] The system calculates the differences between other historical demand feature parameters and the initial matching feature parameters in three dimensions: business type, participating departments, and time requirements. In the business type dimension, if both are completely identical, the difference is 0; if there are differences in sub-types, the difference is calculated based on preset business type difference weights. In the participating departments dimension, the overlap of participating departments is calculated. For example, if the initial matching feature parameter has participating departments in purchasing and finance, while another record has participating departments in purchasing, finance, and warehousing, the overlap is 2 / 3, and the difference is calculated based on this overlap. In the time requirements dimension, the difference in time span is calculated. For example, if the initial matching feature parameter has a time span of 30 days, while another record has a time span of 40 days, the difference is 10 days, which is then converted into a standard difference value based on the importance weight of the time span.
[0085] The sum of the differences across all dimensions is the initial matching cost. For example, if the difference for business type is 0.5, the difference for participating departments is 0.3, and the difference for time requirement is 0.2, the sum is 1.0, meaning the initial matching cost is 1.0.
[0086] After calculating the initial matching cost, the system iteratively selects matching feature parameters. Each time, it reselects a previously unselected historical record from the set of historical demand feature parameters as the new matching feature parameter, repeating the process of calculating the sum of differences to obtain a new matching cost. The system compares the matching costs obtained each time until it finds the target matching feature parameter with the minimum matching cost.
[0087] The system calculates the degree of matching between the current requirement's characteristic parameters and the target's matching characteristic parameters, using this as the requirement matching coefficient. The matching degree is calculated based on similarity across three dimensions: for example, identical business types result in a similarity of 1; higher overlap in participating departments leads to higher similarity; and closer time spans result in higher similarity. A weighted calculation yields the final matching degree, which is used in subsequent process matching calculations to determine the optimal collaborative process template. This entire process, through the analysis and matching of historical data, accurately identifies the requirement characteristics of the current collaborative request, providing precise requirement analysis support for enterprise collaborative management and ensuring efficient matching and execution of subsequent collaborative processes.
[0088] Example 3: In the implementation of the process matching module, the system needs to retrieve and calculate the optimal collaborative process template based on the demand matching coefficient and historical process matching data. The system will obtain the historical process matching error set used for collaborative matching with the enterprise collaborative process template library. This error set contains error data generated by the enterprise in the past when using the collaborative process template library to process various collaborative requests. Specifically, it covers the deviation information between the actual process execution effect and the expected effect of the template in each matching process, such as the deviation of task completion time, the difference in resource consumption, and the inconsistency of the execution order of process nodes. This error data is stored in the database in a structured manner. Each record corresponds to one collaborative matching operation and records in detail the collaborative request information at that time, the selected process template, and the various error values generated during the actual execution.
[0089] The system performs statistical analysis on the historical process matching error set, calculating the average error value of the set, which serves as the baseline error for the process template. Specifically, the system iterates through all records in the error set, sums the error values for each record, and then divides this sum by the total number of records in the error set to obtain an average error value that comprehensively reflects the historical matching error level. For example, if the error set contains 100 records, each with 3 error values, the system first sums each error value for all 100 records, then divides each sum by 100 to obtain the average value of each error. Finally, a weighted calculation method (assigning different weights based on the importance of each error) is used to derive the final baseline error for the process template. This baseline error objectively reflects the overall error level of the enterprise's past collaborative matching using the process template library, providing an important reference benchmark for subsequent matching calculations.
[0090] After obtaining the baseline error of the process template, the system calculates the process matching correction coefficient based on the requirement matching coefficient and the baseline error. The requirement matching coefficient is obtained by the collaborative requirement identification module through matching and analysis of the requirement characteristic parameters of the current collaborative request with those of historical requirements; it reflects the similarity between the current and historical requirements. The calculation logic of the process matching correction coefficient is as follows: a high requirement matching coefficient indicates a high similarity between the current and historical requirements, allowing for minor corrections to the matching process based on the baseline error of historical process matching; conversely, a low requirement matching coefficient indicates that the current requirement has certain unique characteristics, necessitating greater corrections to adapt to the new requirements. The specific calculation method involves substituting the requirement matching coefficient and the process template baseline error into a preset function, and then performing calculations to obtain the process matching correction coefficient. For example, the preset function might be: Process Matching Correction Coefficient = Requirement Matching Coefficient × Process Template Baseline Error × Correction Factor (the correction factor is a constant set according to the company's actual situation). This calculation ensures that the process matching correction coefficient simultaneously reflects the degree of requirement matching and the error level of historical matching.
[0091] Based on the obtained process matching correction coefficients, the system begins to calculate the matching degree between the collaborative request information and each template in the process template library. The process template library stores various collaborative process templates pre-designed by the enterprise for different business types, participating departments, and time requirements. Each template includes detailed information such as process nodes, node execution order, node responsible person, and resource allocation requirements. During the matching degree calculation, the system compares the information of the current collaborative request (including requirement characteristic parameters such as business type, participating departments, and time requirements) with the applicable scenarios and parameters of each template in the template library. For each template, the system calculates its matching degree with the current collaborative request in dimensions such as business type, participating departments, and time requirements. For example, a complete match in business type scores 100 points, an 80% overlap in participating departments scores 80 points, and a time requirement within a preset tolerance range scores 90 points. Then, the system weights and sums these scores to obtain the initial matching degree score between the template and the current collaborative request.
[0092] Based on this, the system adjusts the initial matching score using a process matching correction coefficient to obtain the final matching score. Specifically, if the process matching correction coefficient is greater than 1, the initial matching score needs to be increased to reflect the correction of historical matching errors; if the process matching correction coefficient is less than 1, the initial matching score is decreased. For example, if a template's initial matching score is 85 points and the process matching correction coefficient is 1.2, then the final matching score is 85 × 1.2 = 102 points (if the score cap is 100 points, then 100 points is used). This method ensures that the final matching score more accurately reflects the template's applicability to the current collaborative request, considering both the direct matching degree between the template and the requirement, and correcting for historical matching errors.
[0093] The system sorts all templates in the process template library based on their final matching scores and selects the template with the highest matching score as the optimal collaborative process template. For example, if template A has a final matching score of 95, template B has 92, and template C has 88, the system will select template A as the optimal collaborative process template for the current collaborative request. The selected template will be used in subsequent collaborative execution processes, providing specific process guidance and standards for the collaborative tasks. The entire process matching module, through the analysis and utilization of historical data combined with the characteristics of current needs, can efficiently and accurately match the most suitable process template for collaborative requests, thereby improving the efficiency and accuracy of enterprise collaborative management and ensuring that collaborative tasks can be executed smoothly according to the preset process.
[0094] Example 4: In the implementation of the execution feedback module, the system needs to collect, analyze, and dynamically adjust data on the collaborative execution process based on the optimal collaborative process template. Taking a production planning collaboration scenario of a manufacturing enterprise as an example, after the system starts the collaborative task according to the optimal template selected by the process matching module, the execution feedback module will collect data on task completion progress, resource usage, and time consumption in real time during the collaborative execution process. For example, in production planning collaboration, task completion progress data includes the order completion volume and process compliance rate of each production line; resource usage covers equipment operating time, raw material consumption, and manpower shifts; and time consumption data involves the start and end times of each production link and the interval between process connections. This data is transmitted to the system database in real time through channels such as IoT sensors deployed by the enterprise and production management system interfaces, forming structured execution process data.
[0095] The system needs to train an execution feedback analyzer that includes multi-path feedback analysis using machine learning methods. First, the system collects a sample execution process dataset from historical co-execution records, such as the execution data of all production plan co-task within the past six months, including task progress, resource consumption, etc. under different order scales, equipment status, and raw material supply conditions. At the same time, the feedback adjustment coefficients corresponding to each sample execution process data are collected. These coefficients are records of the adjustment efforts taken for different execution deviations in history. For example, when the progress lags due to equipment failure, the system once increased the manpower input of the subsequent process by 20%, and this 20% is recorded as a feedback adjustment coefficient. All sample feedback adjustment coefficients are classified and labeled according to the adjustment type (such as task progress adjustment, resource allocation adjustment, schedule adjustment) to form a sample feedback adjustment coefficient set.
[0096] The system groups and divides the sample execution process dataset and the sample feedback adjustment coefficient set according to the preset analysis quantity (such as preset to 5 analysis paths). For example, the historical data is divided into 5 groups according to the order scale, and each group contains samples with different equipment status and raw material supply conditions to ensure the diversity of each group of data. For each group of data, the system trains a feedback analysis path based on the supervised learning algorithm, and each path corresponds to different feature extraction methods and prediction logics. For example, the first path focuses on analyzing the correlation between equipment failure and progress deviation, the second path focuses on the impact of raw material shortage on resource allocation, etc. The 5 paths are combined to form an execution feedback analyzer, which can analyze the execution process data from multiple dimensions.
[0097] After obtaining the execution feedback analyzer, the system calculates the actual analysis path quantity according to the demand matching coefficient of the current co-task and the preset analysis quantity. Suppose the current demand matching coefficient is 0.8 and the preset analysis quantity is 5, then the actual analysis path quantity is 0.8×5 = 4 paths. The system randomly selects 4 paths from the 5 feedback analysis paths of the execution feedback analyzer. For example, it randomly selects the equipment failure analysis path, the raw material consumption analysis path, the manpower input analysis path, and the time node analysis path. The collected execution process data (such as the daily output of a production line fails to meet the standard by 20% due to equipment maintenance, the raw material consumption exceeds the budget by 15%, and the process connection time is extended by 30 minutes) is input into these 4 paths, and each path outputs the corresponding feedback adjustment coefficient. For example, the equipment failure path outputs an adjustment coefficient of "increase the debugging manpower of standby equipment by 10%", and the raw material path outputs a coefficient of "initiate an emergency procurement process and shorten the supply cycle by 2 days", etc., to form a feedback adjustment coefficient set containing 4 adjustment coefficients.
[0098] The system calculates the average of this set of coefficients as the final feedback adjustment coefficient. For example, if the four coefficients are 10%, 2 days (converted to a standardized value such as 0.4, assuming 1 day corresponds to 0.2), 5%, and 0.3 (standardized value for time adjustment), the average value after standardization is (0.1 + 0.4 + 0.05 + 0.3) / 4 = 0.2125. Based on this feedback adjustment coefficient, the system dynamically adjusts the collaborative execution process. If the adjustment coefficient is positive, it indicates that resource input needs to be increased or task progress needs to be accelerated. For example, in production planning collaboration, the system automatically increases the number of shifts for the next 3 days by 21.25% and sends an expedited raw material purchase order to the purchasing department. If it is negative, resource input is reduced or the schedule is adjusted, such as reducing overtime shifts due to ahead-of-time progress.
[0099] After adjustments are made, the system continuously collects data on the adjusted execution process, generating execution feedback results. For example, if adjusting manpower and procurement results in an 18% increase in production line output the following day, the risk of raw material shortages is eliminated, and process connection times are shortened to the standard range, these data will be recorded in the feedback results for subsequent collaborative optimization. Throughout the process, the system uses multi-path machine learning analysis and dynamic adjustment mechanisms to achieve real-time monitoring and optimization of the collaborative execution process, ensuring that collaborative tasks can be corrected promptly when deviations occur, closely approximating the expected results of the optimal process template. This implementation method, through a data-driven feedback mechanism, combines historical experience with real-time data, enabling enterprise collaborative management to possess adaptive and dynamic adjustment capabilities, suitable for execution process control in various complex business scenarios.
[0100] Example 5: In the collaborative requirement identification module, the implementation process of extracting the business type, participating departments, and time requirements of collaborative requests from the digital twin model and obtaining the historical requirement feature parameter set can be explained in detail using a collaborative scenario of a retail company's promotional activities. For example, when the company initiates a collaborative request for a "Double 11" promotional activity, the system first parses the collaborative request text, which contains content such as "online and offline promotional activity planning, product preparation, logistics and distribution, and customer service support." The system uses natural language processing technology to identify business keywords such as "promotional activity," "product preparation," and "logistics and distribution." Then, these keywords are matched with a preset business type classification directory. This directory has a "promotional activity management" subcategory under the "marketing" category, thus determining the business type as "marketing - promotional activity management."
[0101] Regarding the extraction of participating departments, the collaboration request text explicitly states that "the Marketing Department will take the lead, with the Merchandising Department, Logistics Department, and Customer Service Department assisting in the execution." The system directly extracts the responsible department as the Marketing Department, and the assisting departments as the Merchandising Department, Logistics Department, and Customer Service Department. Simultaneously, the system confirms this by combining organizational structure data from the digital twin model. This data shows that the Marketing Department is responsible for the planning and execution of the company's marketing activities, the Merchandising Department is responsible for product procurement and inventory management, the Logistics Department undertakes logistics and distribution tasks, and the Customer Service Department is responsible for customer consultation and service. The responsibilities of these departments are consistent with the description in the collaboration request, thus ensuring the accuracy of the participating department information.
[0102] Regarding the handling of time requirements, the collaboration request specifies that "preparatory work for the promotional activity will begin on November 1, 2025, the activity will officially begin on November 11, 2025, and end on November 15, 2025." The system extracts the start date as November 1, 2025, and the completion date as November 15, 2025, calculating a time span of 15 days, which is used as the time requirement. Subsequently, the system combines the business type "Marketing - Promotional Activity Management," the participating departments "Marketing Department, Merchandising Department, Logistics Department, Customer Service Department," and the time span "15 days" to form complete requirement characteristic parameters.
[0103] Next, the system retrieves the historical demand feature parameter set for similar collaborative requests within a historical timeframe. Based on the current collaborative request's business type, "Marketing - Promotional Activity Management," the system filters out all historical request records in the historical database that are labeled with this business type. For example, it filters out 50 historical collaborative request records from the past two years, including "618 Promotional Activity," "Anniversary Promotional Activity," and "Double 12 Promotional Activity." For each record, the system extracts the participating department information and time requirement data. For instance, the participating departments for the "618 Promotional Activity" are Marketing Department, Merchandising Department, and Logistics Department, with a time span of 10 days; the participating departments for the "Anniversary Promotional Activity" are Marketing Department, Merchandising Department, Logistics Department, and Customer Service Department, with a time span of 7 days, etc. This information is then organized according to a unified format to form a historical demand feature parameter set containing three dimensions: business type, participating departments, and time requirements. Each record in this set details the participating departments and time span for different promotional activities.
[0104] When randomly selecting initial matching feature parameters from the historical demand feature parameter set and calculating the initial matching cost, the system randomly selects one record from 50 historical records. Assuming the selected record is "2024 Double 12 Promotion Activity," its business type is "Marketing - Promotion Activity Management," the participating departments are the Marketing Department, Merchandising Department, and Logistics Department, and the time span is 8 days. Using this as the initial matching feature parameter, the system begins to calculate the differences between the other 49 historical records and this initial parameter in terms of business type, participating departments, and time requirements.
[0105] In the business type dimension, since all historical records are of the "Marketing - Promotional Activity Management" business type, the difference value is 0. In the participating department dimension, the overlap of participating departments is calculated. For example, if a historical record involves the Marketing Department, Merchandising Department, Logistics Department, and Customer Service Department, compared to the initial parameters (Marketing Department, Merchandising Department, Logistics Department), the overlap is 3 / 4 = 0.75, and the difference value is 1 - 0.75 = 0.25. In the time requirement dimension, the difference in time span is calculated. The initial parameter's time span is 8 days, while a historical record's time span is 12 days, resulting in a difference of 4 days. According to the preset time span difference conversion rule (e.g., a difference value of 0.1 for every day difference), this dimension's difference value is 0.4. Adding the difference values of each dimension, the total difference between this historical record and the initial matching feature parameters is 0 + 0.25 + 0.4 = 0.65, which is used as part of the initial matching cost.
[0106] The system iteratively selects matching feature parameters. Each time, it randomly selects a previously unselected historical record from the set of historical demand feature parameters as the new matching feature parameter, repeating the process of calculating the sum of differences to obtain a new matching cost. For example, the second time the record "2024 Anniversary Promotion" is selected, its participating departments are Marketing, Merchandising, and Logistics, and its time span is 7 days. The sum of differences with other historical records is calculated to obtain a new matching cost. The system compares the matching costs obtained each time until it finds the target matching feature parameter with the minimum matching cost. Suppose that after multiple iterations, the record "2023 Double 11 Promotion" is found to have the smallest sum of differences with the current demand feature parameter, at 0.32; this record is the target matching feature parameter.
[0107] Finally, the system calculates the degree of matching between the current demand feature parameters and the target matching feature parameters, which serves as the demand matching coefficient. In the business type dimension, the matching degree is 1. In the participating department dimension, the participating departments for the current demand are Marketing, Merchandising, Logistics, and Customer Service, while the participating departments for the target matching feature parameters are Marketing, Merchandising, Logistics, and Customer Service (assuming the participating departments in the historical records are the same as the current ones), resulting in an overlap of 1 and a matching degree of 1. In the time requirement dimension, the current time span is 15 days, while the time span for the target matching feature parameters is 14 days. According to the preset time span matching degree calculation rules (e.g., a 1-day difference in time span results in a matching degree of 0.95), the matching degree for this dimension is 0.95. Through weighted calculation (assuming the weights for business type, participating departments, and time requirement are 0.4, 0.4, and 0.2 respectively), the matching degree is calculated as 1×0.4 + 1×0.4 + 0.95×0.2 = 0.4 + 0.4 + 0.19 = 0.99, meaning the demand matching coefficient is 0.99. This coefficient indicates that the current collaborative request's demand characteristics are highly similar to those of the historical "2023 Double 11 promotional event," providing an accurate basis for subsequent process matching. The system can use this coefficient to retrieve a more suitable collaborative process template, ensuring the smooth planning and execution of collaborative tasks for the promotional event. The entire process, through specific data processing of a retail enterprise's promotional event collaboration scenario, demonstrates the complete implementation path from demand characteristic extraction and historical data matching to finally obtaining the demand matching coefficient, reflecting the system's application logic and operational details in a real-world business scenario.
[0108] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A digital twin enterprise collaborative management platform system, characterized in that, The system includes: The digital twin building module is used to collect basic enterprise data, establish a multi-dimensional mapping relationship between physical and virtual objects, and generate an enterprise digital twin model; The collaborative demand identification module is used to obtain enterprise collaborative request information based on the digital twin model, extract demand feature parameters, and perform feature matching analysis in combination with historical collaborative demand data to obtain demand matching coefficients. The process matching module is used to retrieve the enterprise collaborative process template library based on the required matching coefficient, and use the process matching feature set in the historical process matching data to perform process matching calculation on the collaborative request information to obtain the optimal collaborative process template, including: Obtain the historical process matching error set that uses the enterprise collaborative process template library for collaborative matching; Calculate the average error value of the historical process matching error set as the baseline error for the process template; Based on the demand matching coefficient and the process template baseline error, the process matching correction coefficient is calculated. Based on the process matching correction coefficient, the matching degree of the collaboration request information and each template in the process template library is calculated, and the template with the highest matching degree is selected as the optimal collaboration process template. The execution feedback module is used to collect collaborative execution process data based on the optimal collaborative process template, analyze the execution deviation magnitude, calculate the feedback adjustment coefficient, dynamically adjust the collaborative execution process, and generate execution feedback results, including: Collect data on task completion progress, resource usage, and time consumption during collaborative execution as execution process data; A machine learning method is used to train an execution feedback analyzer that includes multi-path feedback analysis, wherein the number of paths in the multi-path feedback analysis is a preset number of analyses. Multiply the demand matching coefficient by the preset analysis quantity to obtain the actual number of analysis paths; Within the execution feedback analyzer, a feedback analysis path of the actual number of analysis paths is randomly selected, and the execution process data is input to obtain a set of feedback adjustment coefficients. Calculate the average value of the set of feedback adjustment coefficients, and use it as the feedback adjustment coefficient; Based on the feedback adjustment coefficient, the task progress, resource allocation, and time arrangement during the collaborative execution process are adjusted positively or negatively to form execution feedback results; A machine learning approach is used to train an execution feedback analyzer that incorporates multi-path feedback analysis, including: Based on historical collaborative execution records, a dataset of sample execution processes was collected, and feedback adjustment coefficients under different sample execution process data were collected and labeled as a set of sample feedback adjustment coefficients. According to the preset number of analyses, the sample execution process dataset and the sample feedback adjustment coefficient set are grouped and divided to obtain multiple sets of feedback training data; Using the multiple sets of feedback training data respectively, a preset number of feedback analysis paths are trained based on a supervised learning algorithm, and then combined to form an execution feedback analyzer.
2. The digital twin enterprise collaborative management platform system according to claim 1, characterized in that, Collect basic enterprise data, establish a multi-dimensional mapping relationship between physical and virtual objects, and generate a digital twin model of the enterprise, including: Acquire enterprise operation data, organizational structure data, and resource allocation data, and process them into standardized formats respectively; The processed operational data is mapped to the virtual operational scenario, the organizational structure data is mapped to the virtual organizational nodes, and the resource configuration data is mapped to the virtual resource units. Integrate all related mapping relationships to generate a digital twin model of the enterprise covering operational, organizational, and resource dimensions.
3. The digital twin enterprise collaborative management platform system according to claim 1, characterized in that, Based on the digital twin model, enterprise collaboration request information is obtained, demand feature parameters are extracted, and feature matching analysis is performed in conjunction with historical collaboration demand data to obtain demand matching coefficients, including: Extract the business type, participating departments, and time requirements involved in the collaborative request from the digital twin model as demand feature parameters; Obtain the set of historical demand characteristic parameters for similar collaborative requests within a historical time range; Randomly select initial matching feature parameters from the set of historical demand feature parameters, and calculate the sum of the differences between the initial matching feature parameters and other historical demand feature parameters to obtain the initial matching cost; The matching feature parameters are selected iteratively until the target matching feature parameter with the minimum matching cost is found. Calculate the degree of matching between the current demand feature parameters and the target matching feature parameters, and use it as the demand matching coefficient.
4. The digital twin enterprise collaborative management platform system according to claim 3, characterized in that, Randomly select initial matching feature parameters from the set of historical demand feature parameters, and calculate the sum of the differences between the initial matching feature parameters and other historical demand feature parameters to obtain the initial matching cost, including: Randomly select one historical demand feature parameter from the set of historical demand feature parameters as the initial matching feature parameter; Calculate the differences between other historical demand feature parameters and the initial matching feature parameters in terms of business type, participating department, and time requirement, and add the differences in each dimension to obtain the total difference magnitude; The sum of the difference magnitudes is used as the initial matching cost.
5. The digital twin enterprise collaborative management platform system according to claim 2, characterized in that, Acquire enterprise operational data, organizational structure data, and resource allocation data, and process them into standardized formats, including: Real-time collection of business transaction records, production progress reports, and customer service logs from operational data via enterprise information system interfaces; Extract departmental hierarchy, job information, and personnel affiliation from organizational structure data through the human resource management system; The asset management system is used to obtain data on equipment distribution locations, material inventory quantities, and cash flow records from resource allocation data. The collected operational data, organizational structure data, and resource allocation data are uniformly converted into structured tables, and duplicates and invalid fields are removed to complete the standardized format processing.
6. The digital twin enterprise collaborative management platform system according to claim 3, characterized in that, The business type, participating departments, and time requirements involved in the collaborative request are extracted from the digital twin model as demand feature parameters, including: Parse the collaborative request text content, identify the business keywords within it, match it with the preset business type category directory, and determine the business type; Extract the names of the responsible and cooperating departments explicitly mentioned in the collaboration request, and confirm the participating departments by combining them with the organizational structure data in the digital twin model; Extract the start and finish time nodes specified in the collaboration request, calculate the time span, and use it as the time requirement; The business type, participating departments, and time span are combined to form the demand characteristic parameters.
7. The digital twin enterprise collaborative management platform system according to claim 3, characterized in that, Obtain a set of historical demand feature parameters for similar collaborative requests within a historical time range, including: filtering historical request records with the same business type as the current collaborative request, extracting the participating department information and time requirement data from each record, and organizing them into a set of historical demand feature parameters containing business type, participating department, and time requirement.