Intelligent progress management and control method and system for information engineering supervision

Abstract

The invention provides an intelligent progress management and control method and system for information engineering supervision, and relates to the technical field of information, and the method comprises the steps: pushing early warning information to a first-party system and a second-party system through message middleware, collecting a feedback state to complete closed-loop cooperative processing, and updating a processing result to a central database; based on the central database, combining historical project data, performing pattern recognition analysis to extract a progress deviation rule, and dynamically optimizing a plan node threshold and buffer time of a subsequent stage according to the progress deviation rule to obtain an optimization strategy; and on the basis of the optimization strategy, constructing a configurable progress management and control template library, so as to quickly adapt and generate a corresponding progress management and control node and an early warning threshold value in the new-type project. According to the invention, a full-process closed-loop management system from data integration, intelligent optimization to dynamic adaptation is constructed, and the fundamental transformation of project progress management and control from passive response to active prediction and from experience driving to data driving is realized.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to an intelligent progress control method and system for information engineering supervision. Background Technology

[0002] As enterprise informatization construction deepens, the scale of large-scale information engineering projects (such as ERP system upgrades and data center construction) continues to expand, and the technical complexity continues to increase. These projects typically involve multiple parties, including clients, vendors, and third-party testing organizations. The use of heterogeneous management systems (such as OA, Jira, TestRail, etc.) by these parties for project management may result in dispersed data sources and inconsistent formats.

[0003] In practical applications, taking the ERP system upgrade project of a manufacturing group as an example, the project cycle was 6 months, involving multiple stages such as requirements research, system development, module testing, data migration, and trial operation. Under the traditional supervision method, data from all parties was stored in isolation and lacked an automatic synchronization mechanism. Supervisors needed to manually log in to each platform every day to extract data and perform secondary entry, which may result in a data delay of 1 to 2 days. At the same time, manually filled Excel spreadsheets had problems such as formula errors and missing fields, and data cleaning relied entirely on the supervisor's experience.

[0004] Existing data processing methods have the following drawbacks: First, data collection is fragmented, forming information silos and lacking an effective automatic synchronization mechanism; second, data quality control mechanisms are lacking, and erroneous data may affect the accuracy of supervisory decisions; third, data processing is disconnected from business control, which may prevent the identification of the transmission effect of critical path deviations; and fourth, the value of historical data is not fully utilized, which may prevent the prediction of risks through historical patterns. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent progress control method and system for information engineering supervision. The method and system construct a closed-loop management system for the entire process from data integration, intelligent optimization to dynamic adaptation, and realize the fundamental transformation of project progress control from passive response to proactive prediction and from experience-driven to data-driven.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, an intelligent progress control method for information engineering supervision, the method comprising: Collect raw progress data from the testing systems of Party A, Party B, and third parties, perform data cleaning and format standardization to obtain regularized data, and store it in the central database; Based on the regularized data, multiple quality control nodes are set in the data processing flow, and a multi-dimensional feature space is constructed based on the data processing quality indicators of the quality control nodes; the multi-dimensional feature space is divided into regions, and the data feature distribution patterns of each region are identified to obtain data processing optimization parameters; the data processing process is dynamically optimized using the optimization parameters to obtain optimized real-time data. The optimized real-time data is automatically compared with the preset key nodes and thresholds of the project plan to obtain the comparison results. Based on the comparison results, real-time early warning judgment is made. When a deviation from the schedule or failure to meet the key indicators is identified, an early warning event is automatically generated, and the transmission impact on the critical path of the project is analyzed through a pre-configured rule engine to obtain early warning information. The warning information is pushed to the systems of Party A and Party B through the message middleware and feedback status is collected to complete the closed-loop collaborative processing. At the same time, the processing results are updated to the central database. Based on the central database and combined with historical project data, pattern recognition analysis is performed to extract the pattern of progress deviation, and the planned node thresholds and buffer times of subsequent stages are dynamically optimized accordingly to obtain optimization strategies. Based on optimization strategies, a configurable schedule management template library is built to quickly adapt to new types of projects and generate corresponding schedule management nodes and early warning thresholds.

[0007] Secondly, an intelligent progress control system for information engineering supervision includes: The data acquisition module is used to collect raw progress data from the testing systems of the client, vendor, and third parties, perform data cleaning and format standardization to obtain regularized data, and store it in the central database. The dynamic optimization module is used to set multiple quality control nodes in the data processing flow based on the data stored in the central database, and to construct a multi-dimensional feature space based on the data processing quality indicators of the quality control nodes; to divide the multi-dimensional feature space into regions, identify the data feature distribution patterns of each region, and obtain data processing optimization parameters; and to dynamically optimize the data processing process using the optimization parameters to obtain optimized real-time data. The comparison module is used to automatically compare the optimized real-time data with preset key nodes and thresholds of the project plan to obtain the comparison results; The early warning module is used to make real-time early warning judgments based on comparison results. When a deviation from the schedule or failure to meet key indicators is identified, an early warning event is automatically generated, and its transmission impact on the critical path of the project is analyzed through a pre-configured rule engine to obtain early warning information. The processing module is used to push the early warning information to the systems of Party A and Party B through the message middleware and collect feedback status to complete the closed-loop collaborative processing, while updating the processing results to the central database. The optimization strategy module is used to perform pattern recognition analysis based on the central database and historical project data to extract the pattern of progress deviation, and dynamically optimize the planned node thresholds and buffer times of subsequent stages to obtain the optimization strategy. The adaptation module is used to build a configurable progress management template library based on optimization strategies, so as to quickly adapt to new types of projects and generate corresponding progress management nodes and early warning thresholds.

[0008] Thirdly, a computing device includes: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0009] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0010] The above-described solution of the present invention has at least the following beneficial effects: This system collects raw progress data from the client, vendors, and third-party testing systems, integrating scattered data from different heterogeneous management systems. Data cleaning and format standardization eliminate format differences, resolving issues of scattered data sources and inconsistent formats. Storing the standardized data in a central database enables centralized data storage, facilitating rapid data retrieval in subsequent stages. Multiple quality control nodes are set up in the data processing flow to control quality at critical stages, reducing the entry of erroneous data into later processes. A multi-dimensional feature space is constructed based on the data processing quality indicators of the quality control nodes, comprehensively characterizing the multi-dimensional features of data processing quality. Regional division of the multi-dimensional feature space and identification of data feature distribution patterns reveal potential problems and optimization directions in the data processing process. Dynamic optimization of the data processing process using optimization parameters allows adjustments based on actual conditions, continuously improving data processing effectiveness and obtaining high-quality optimized real-time data. The optimized real-time data is automatically compared with preset project plan key nodes and thresholds, reducing manual operations. Automatic comparison ensures consistency and quickly obtains comparison results, providing timely and accurate basis for subsequent early warning judgments and improving the efficiency of data processing and project management.

[0011] Real-time early warning based on comparison results can promptly identify progress deviations or failures to meet key performance indicators. By analyzing the transmission impact of early warning events on the project's critical path through a pre-configured rule engine, the system can systematically analyze the scope of the problem's influence on core project processes, ensuring that early warning information not only includes the problem itself but also its impact on key project stages, providing a more comprehensive reference for subsequent handling. Pushing early warning information to both client and vendor systems via message middleware enables efficient transmission of early warning information across multiple systems, ensuring timely access to warning content for all parties. Collecting feedback status and completing closed-loop collaborative processing allows for effective linkage among all parties regarding the handling of early warning events. Updating processing results to the central database ensures the timeliness and completeness of the central database data. Pattern recognition analysis based on the central database and historical project data can fully utilize historical data... The value of historical data lies in uncovering common patterns in project schedule deviations. Extracting these patterns provides data support for subsequent plan adjustments. Based on this, the thresholds and buffer times for planned nodes in later stages can be dynamically optimized, making plan parameters more aligned with actual project execution, improving the rationality and feasibility of the plan, and yielding optimized strategies adapted to project progress. A configurable schedule management template library can be built based on these optimization strategies, transforming optimization experience into reusable template resources and reducing repetitive work in configuring management for new projects. By extracting characteristic parameters from new project types and matching them with baseline templates in the template library, suitable basic templates can be quickly determined. Customized schedule management nodes are generated through parameter adaptation adjustments, and warning thresholds are generated using a rule engine. This allows new projects to quickly have a schedule management system tailored to their needs, improving the efficiency and adaptability of new project schedule management. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating an intelligent progress control method for information engineering supervision provided by an embodiment of the present invention.

[0013] Figure 2 This is a schematic diagram of an intelligent progress control system for information engineering supervision provided by an embodiment of the present invention. Detailed Implementation

[0014] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0015] like Figure 1 As shown, an embodiment of the present invention proposes an intelligent progress control method for information engineering supervision, the method comprising the following steps: Step 100: Collect the original progress data from the testing systems of Party A, Party B, and third parties, perform data cleaning and format standardization processing to obtain regularized data, and store it in the central database. Step 200: Based on the regularized data, multiple quality control nodes are set in the data processing flow, and a multi-dimensional feature space is constructed based on the data processing quality indicators of the quality control nodes; the multi-dimensional feature space is divided into regions, and the data feature distribution patterns of each region are identified to obtain data processing optimization parameters; the data processing process is dynamically optimized using the optimization parameters to obtain optimized real-time data. Step 300: The optimized real-time data is automatically compared with the preset key nodes and thresholds of the project plan to obtain the comparison results; Step 400: Based on the comparison results, perform real-time early warning judgment. When a deviation from the schedule or failure to meet key indicators is identified, an early warning event is automatically generated, and its transmission impact on the critical path of the project is analyzed through a pre-configured rule engine to obtain early warning information. Step 500: Push the warning information to the systems of Party A and Party B through the message middleware and collect feedback status to complete the closed-loop collaborative processing, and update the processing results to the central database at the same time. Step 600: Based on the central database and combined with historical project data, perform pattern recognition analysis to extract the schedule deviation pattern, and dynamically optimize the planned node thresholds and buffer times for subsequent stages to obtain the optimization strategy. Step 700: Based on the optimization strategy, build a configurable progress control template library to quickly adapt to and generate corresponding progress control nodes and early warning thresholds in new types of projects.

[0016] In this embodiment of the invention, multi-source raw progress data from the client, contractor, and third-party testing systems are integrated. Abnormal information is removed through data cleaning, and the presentation format of data from different sources is standardized to obtain well-organized and highly consistent data. This well-organized data is then stored in a central database for centralized management, providing a high-quality and easily accessible data foundation for subsequent data processing and analysis. Multiple quality control nodes are set up in the data processing flow to control the quality of each stage of data processing and reduce quality problems. A multi-dimensional feature space is constructed based on the data processing quality indicators of the quality control nodes, which can comprehensively characterize the data processing quality-related features. The multi-dimensional feature space is then further processed... By dividing the data into regions and identifying the distribution patterns of data characteristics in each region, the inherent characteristics of the data processing process can be accurately mined. The resulting data processing optimization parameters can be dynamically adjusted to improve the adaptability of the data processing process and the quality of data output. The optimized real-time data is automatically compared with preset key nodes and thresholds of the project plan, reducing manual operation and improving the efficiency of the comparison work. Based on the comparison results, real-time early warning judgments can be automatically completed, quickly identifying progress deviations or key indicators that fail to meet standards and automatically generating early warning events, avoiding the delays of manual judgment. Through the analysis of early warning events by a pre-configured rule engine, the transmission impact on the critical path of the project can be systematically sorted out.

[0017] By pushing early warning information through message middleware, both the client and the vendor can receive the warning information in a timely manner; collecting feedback status forms a closed-loop collaborative processing mechanism, making the handling process of early warning events traceable and accountable; updating the processing results to the central database ensures the up-to-dateness of the data in the central database, providing accurate data for subsequent data analysis and strategy optimization; relying on historical project data in the central database, combined with pattern recognition analysis, it is possible to deeply explore potential patterns related to project progress and effectively extract schedule deviation patterns; based on the extracted schedule deviation patterns, the planned node thresholds and buffer times of subsequent stages are dynamically optimized, making the planning settings of subsequent stages more in line with the actual situation of the project, improving the feasibility and rationality of the plan; based on the optimization strategy, a configurable schedule control template library can be built to achieve standardized storage of core elements of schedule control; in new types of projects, the template library can be used to quickly adapt to project requirements, generate corresponding schedule control nodes and early warning thresholds, eliminate the need to design control schemes from scratch, reduce repetitive workload, and improve the efficiency and adaptability of schedule control schemes for new types of projects.

[0018] In a preferred embodiment of the present invention, step 100 above involves collecting raw progress data from the testing systems of Party A, Party B, and third parties, performing data cleaning and format standardization processing to obtain regularized data, and storing it in a central database, including: Step 101: Establish connections with the client's management system, the vendor's development system, and the third-party testing system through multiple pre-configured data interfaces, and collect raw progress data from these systems in real time. Specifically, this includes: pre-configuring adapted data interfaces for each of the client's management system (e.g., an OA system), where the interface adapted to the client's management system (e.g., a Jira system) supports reading data from the project progress reporting module; the interface adapted to the vendor's development system (e.g., a Jira system) supports extracting the completion status and progress node information of development tasks; and the interface adapted to the third-party testing system (e.g., a TestRail system) supports obtaining test case pass rates and defect repair progress data. Establish stable connections between each pre-configured interface and the corresponding system's database or API service using an interface configuration tool, setting encryption protocols for data transmission during configuration to ensure data security. After the connection is established, start the real-time data collection task, setting it to incrementally capture target data from each system every 5 minutes, automatically filtering duplicate data during the capture process, and transmitting the captured raw progress data to a temporary data cache in real time.

[0019] Step 102 involves cleaning the original progress data by filling in missing data points and correcting abnormal data values. Specifically, this includes: retrieving the collected original progress data from a temporary data cache; first, starting a missing data detection program to verify key fields in the data one by one, identifying fields with missing values, such as project phase name, progress completion percentage, and data collection time; for missing numerical data, such as progress completion percentage, filling in the missing data by using the average of the previous three collected data points for the same project phase; for missing textual data, such as project phase name, obtaining the missing data from the corresponding system by matching the associated task number; then starting an abnormal data identification program, setting a reasonable data range threshold, such as a reasonable range of 0 to 100% for progress completion percentage, and marking data exceeding the threshold as abnormal data; for marked abnormal data, correcting it by comparing the corresponding original system record with the project plan document; if it is confirmed to be a data entry error, adjusting according to the original system record; if it is confirmed to be a temporary data fluctuation, retaining the original data and indicating the reason for the fluctuation; after completing the missing data filling and abnormal data correction, the cleaned data is generated.

[0020] Step 103: Based on the cleaned data, the heterogeneous data from different sources are converted according to a unified structured format to obtain regularized data, and the regularized data is stored in the central database. Specifically, this includes: pre-defining a unified structured data format, which includes project basic information fields, progress data fields, and data source fields. The project basic information fields are project number, project name, and current stage; the progress data fields are task number, task name, completion progress, planned completion time, and actual completion time; and the data source fields are the name of the system to which the data belongs and the collection time. The data type and field length limits for each field are also determined. For example, the project number is a string, the completion progress is a string, and so on. The data type and collection time are timestamps. A format conversion program is initiated to map the cleaned data to fields according to a preset structured format. Specifically, the project progress report values ​​from the client's management system are mapped to the completion progress field, and the task status from the vendor's development system is converted to the corresponding completion progress value (e.g., 100% for completed, 50% for in progress). The test pass rate from the third-party testing system directly corresponds to the completion progress field. After format conversion, standardized data is generated. A data import program is then initiated to classify and store the standardized data in the central database according to project number and collection time. Simultaneously, indexes for project numbers and data fields are created in the database to facilitate subsequent data queries and retrieval, achieving unified storage and management of heterogeneous data from different sources.

[0021] In a preferred embodiment of the present invention, step 200 involves setting multiple quality control nodes in the data processing flow based on the regularized data, and constructing a multi-dimensional feature space based on the data processing quality indicators of the quality control nodes; dividing the multi-dimensional feature space into regions, identifying the data feature distribution patterns of each region, and obtaining data processing optimization parameters; and dynamically optimizing the data processing process using the optimization parameters to obtain optimized real-time data, including: Step 201: Based on the standardized data, an input verification node, a transformation logic node, and an output calibration node quality control node are set up in the data processing flow. Specifically, this includes: An input verification node is set at the entry point of the data into the subsequent analysis stage, based on the standardized data processing flow. This node receives the standardized data retrieved from the central database and preliminarily checks whether the data contains basic key fields such as project number, progress completion percentage, and data collection time. A transformation logic node is set up in the intermediate transformation stage of the data processing flow. This node connects the data stream after format standardization and is responsible for monitoring the process of converting heterogeneous data to a unified structured format. An output calibration node is set up at the end of the data processing flow, before the data is output to the progress comparison stage. This node is used to connect to the preset key node data of the project plan in subsequent steps to ensure that the output data can meet the subsequent comparison requirements. The three nodes cover the entire process of data processing—input, intermediate transformation, and output—forming a coherent quality control link.

[0022] Step 202: Based on the quality control node, at the input verification node, count the number of valid records and calculate their proportion to the total number of records to obtain the data integrity ratio. At the conversion logic node, count the number of data entries that conform to the predetermined format specification and calculate their specification conformance ratio to obtain the format specification degree. At the output calibration node, calculate the matching accuracy between the output value and the standard reference value to obtain the numerical accuracy. Specifically, this includes: for the input verification node, retrieving a batch of standardized data, counting the number of records containing the project number, progress completion percentage, data collection time, and no null values ​​as the number of valid records, counting the total number of records in this batch of standardized data, and obtaining the data integrity ratio by dividing the number of valid records by the total number of records; for... For the conversion logic node, pre-set data format specifications, including numeric fields retaining two decimal places, date format being uniformly YYYY-MM-DD, and text fields having no special characters. Count the number of data entries in the data stream processed by this node that conform to the above specifications, and obtain the format conformity by dividing the number of data entries that conform to the specifications by the total number of data entries processed by this node. For the output calibration node, extract the progress standard reference values ​​for each stage from the project plan document provided by the client. For example, the standard progress value for the requirements research stage is 30%, and for the system development stage it is 60%. Calculate the ratio of the absolute deviation of the real-time progress value output by this node from the corresponding standard reference value to the standard reference value, and obtain the numerical accuracy by calculating (1 - absolute deviation ratio).

[0023] Step 203: Based on the data integrity ratio, format standardization, and numerical accuracy, construct a multi-dimensional feature space. Specifically, this includes: constructing a three-dimensional multi-dimensional feature space with the data integrity ratio as the X-axis, format standardization as the Y-axis, and numerical accuracy as the Z-axis, where the value range of the three coordinate axes is set to 0 to 1, and each coordinate axis is divided into scales at intervals of 0.01; using the data integrity ratio, format standardization, and numerical accuracy calculated in step 202 for each batch of data as a set of coordinate values. For example, if the three indicators of a certain batch of data are 0.92, 0.88, and 0.95, then the data point with coordinates (0.92, 0.88, 0.95) is marked in the multi-dimensional feature space. As the batch of data is continuously processed, corresponding data points are continuously added to the multi-dimensional feature space to gradually construct a complete multi-dimensional feature space.

[0024] Step 204 involves calculating the distribution density of each data point in the multidimensional feature space and identifying high-density core regions and low-density edge regions based on a preset density threshold. Specifically, this includes: using a density clustering algorithm to calculate the distribution density of each data point in the multidimensional feature space; setting calculation parameters of a neighborhood radius of 0.05 and a minimum number of data points within the neighborhood of 5; traversing each data point in the multidimensional feature space and counting the number of other data points contained within the neighborhood radius of each data point; classifying target data points with a neighborhood number greater than or equal to 5 and all data points within their neighborhood into the same region, marking them as high-density core regions; and classifying data points with a neighborhood number less than 5 into low-density edge regions, where the minimum number of data points within the neighborhood of 5 is the preset density threshold. This method completes the identification of different density regions in the multidimensional feature space.

[0025] Step 205: Based on the high-density core region and the low-density edge region, perform quantitative analysis on the density gradient change between adjacent regions, and select the points where the gradient change rate reaches the peak to form a boundary point set. Specifically, this includes: selecting the boundary region adjacent to the high-density core region and the low-density edge region as the analysis object, calculating the density difference between two adjacent data points in this region, and simultaneously calculating the spatial distance between these two data points in the multidimensional feature space. The density gradient change rate between the two data points is obtained by dividing the density difference by the spatial distance. Traverse all adjacent data point pairs in the boundary region, record the density gradient change rate of each pair of data points, and select the value with the largest gradient change rate as the peak value. Include the boundary side data points of the two data points corresponding to the peak value into the boundary point set, and continue traversing until all boundary points where the gradient change rate reaches the peak value are collected to form a complete boundary point set.

[0026] Step 206: Based on the set of boundary points, perform topological connection processing on each boundary point to form the boundary range of each region. Specifically, this includes: sorting all boundary points in the set by coordinates, initially sorting the boundary points according to the order of their X-axis coordinates from smallest to largest, and further sorting boundary points with the same X-axis coordinates according to the order of their Y-axis coordinates from smallest to largest; using the sorted boundary points as a basis, employing a topological connection algorithm to connect each boundary point with its two nearest adjacent boundary points, where the first boundary point is also connected to the last boundary point, forming a closed polygon structure; the spatial range enclosed by this closed polygon structure is the boundary range between the high-density core region and the low-density edge region, and the boundary ranges of other regions are constructed in the same way.

[0027] Step 207: Based on the boundary range of each region, perform statistical analysis on the distribution of data points within each boundary range to obtain the data characteristic distribution pattern of each region. Specifically, this includes: for high-density core regions, extracting the values ​​of three indicators—data integrity ratio, format normalization, and numerical accuracy—for all data points within the region, calculating the mean, median, and standard deviation of each indicator, and determining the concentrated distribution range of each indicator. For example, the mean value of the data integrity ratio is 0.93, the median is 0.92, and the standard deviation is 0.02, with a concentrated distribution range of 0.91 to 0.95. For low-density edge regions, statistically analyze the frequency of occurrence of each indicator in different numerical ranges within the region. For example, the data integrity ratio occurs 28 times in the range of 0.7 to 0.8, accounting for 62% of the total number of data points in the region, and the format normalization occurs 25 times in the range of 0.75 to 0.85, accounting for 56% of the total number of data points in the region. The data characteristic distribution pattern of each region is obtained through the above statistical analysis.

[0028] Step 208: Based on the data feature distribution pattern, determine the data aggregation center location of the high-density core area, calculate the coordinate difference between the aggregation center location and the preset ideal centroid location, and obtain the offset vector. Specifically, this includes: based on the data feature distribution pattern of the high-density core area obtained in step 207, calculating the average of three indicators for all data points in the area, and using these three average values ​​as the coordinates of the data aggregation center location of the high-density core area. For example, if the average data integrity ratio is 0.93, the average format standardization is 0.87, and the average numerical accuracy is 0.94, the aggregation center location coordinates are (0.93, 0.87, 0.94). The preset ideal centroid position is set according to the optimal quality requirements of information engineering supervision, such as data integrity ratio 0.98, format standardization 0.95, and numerical accuracy 0.98. The ideal centroid position coordinates are (0.98, 0.95, 0.98). The differences between the aggregation center position coordinates and the ideal centroid position coordinates on the X-axis, Y-axis, and Z-axis are calculated, namely, the difference on the X-axis is 0.93-0.98=-0.05, the difference on the Y-axis is 0.87-0.95=-0.08, and the difference on the Z-axis is 0.94-0.98=-0.04. The vector formed by these three differences (-0.05, -0.08, -0.04) is used as the offset vector.

[0029] Step 209: Based on the offset vector, determine its spatial distance magnitude and angular direction relative to the coordinate axes. Use the magnitude and direction as optimization parameters for the data processing. Specifically, this includes: based on the offset vector (-0.05, -0.08, -0.04) obtained in step 208, calculate the spatial distance magnitude of the vector according to the three-dimensional vector magnitude calculation formula, that is, sum the squares of each coordinate difference in the vector to obtain (-0.05). 2 + (-0.08) 2 + (-0.04) 2 =0.0025+0.0064+0.0016=0.0105, then take the square root of the summation result to obtain the spatial distance modulus; calculate the angle between the offset vector and the positive directions of the X-axis, Y-axis and Z-axis of the multidimensional feature space, and calculate the ratio of the projection of the vector on the three coordinate axes to the vector modulus using the inverse cosine function to obtain the angle value of each coordinate axis direction. The calculated spatial distance modulus and the angle values ​​of the three coordinate axis directions are used together as optimization parameters for the data processing process.

[0030] Step 210: Using the optimization parameters, obtain the adjustment weights of the data cleaning rules and the fault tolerance thresholds for format conversion. Specifically, this includes: According to the optimization parameters, if the spatial distance modulus is large, it indicates that the current data processing quality deviates significantly from the ideal state. Accordingly, increase the adjustment weights of key fields in the data cleaning rules. For example, increase the missing fill weights of key fields such as project number and progress completion percentage from 0.3 to 0.5 to ensure the data integrity of key fields. According to the angular direction of the offset vector on each coordinate axis, if the angular deviation in the Y-axis (format standardization) direction is the largest, it indicates that the quality problem in the format conversion stage is the most prominent. Accordingly, adjust the fault tolerance thresholds for format conversion. For example, extend the fault tolerance range for date format from ±1 day to ±2 days, and extend the precision fault tolerance range for numeric fields from ±0.01 to ±0.02. The adjustment weights of the data cleaning rules and the fault tolerance thresholds for format conversion are obtained through the above methods.

[0031] Step 211: Based on the adjusted weights and fault tolerance thresholds, configure parameters for subsequent data processing to form an optimized data processing flow. Execute the optimized data processing flow to obtain optimized real-time data. Specifically, this includes: applying the adjusted weights obtained in step 210 to the data cleaning stage; prioritizing missing value filling and outlier correction for high-weight fields during subsequent data processing, such as filling missing data in the project number field by retrieving related task data from the vendor's development system; applying the fault tolerance threshold to the format conversion stage; automatically correcting the format of data close to the standard format, such as automatically converting YYYY / MM / DD format date data to YYYY-MM-DD format, and automatically padding numeric data with one decimal place to two decimal places; integrating the above configuration into the data processing flow to form the optimized data processing flow; when new data enters the flow, execute the optimized flow for data processing to ultimately obtain optimized real-time data that meets quality requirements.

[0032] In a preferred embodiment of the present invention, step 300 above, which involves automatically comparing the optimized real-time data with preset key nodes and thresholds of the project plan to obtain the comparison result, includes: Step 301: Parse the preset project plan, extract the key node definitions and their corresponding time and quality thresholds, and establish a project plan structure model. Specifically, this includes: First, obtaining the ERP system upgrade project plan provided by the client, which contains detailed plans for five phases within a 6-month period: requirements gathering, system development, module testing, data migration, and trial operation. Next, begin training the project plan structure model by calling historical data from similar ERP system upgrade projects stored in the central database, such as the key node time and quality threshold records for similar-sized ERP projects in the past 3 years. Extract data such as the historical average time to complete requirements document review (29 days), the historical average time to complete core module coding (88 days), the historical average pass rate for test cases (95.2%), and the historical average pass rate for data migration integrity (99.7%), as references for calibrating the current model thresholds. Then, use a document parsing tool to perform structured parsing of the current project plan, extracting the key node definitions for each phase, such as... The key milestones for the requirements gathering phase are: completion of requirements document review; completion of core module coding; achieving the required pass rate for test cases; completion of historical data migration; and 72 hours of stable system operation. The plan also extracts initial time and quality thresholds for each key milestone. The initial time thresholds include: a planned completion date of requirements document review (30 days after project start); a planned completion date of core module coding (90 days after project start); a planned completion date of the required pass rate for test cases (150 days after project start); a planned completion date of historical data migration (180 days after project start); and a planned date of 72 hours of stable system operation (210 days after project start). The initial quality thresholds include: a requirements document review pass rate of no less than 90%; a core module coding pass rate of no less than 98%; a test case pass rate of no less than 95%; historical data migration integrity of no less than 99.8%; and a system trial operation failure rate of no more than 0.1%.

[0033] Based on the aforementioned historical data, model training (threshold calibration) is conducted. The deviation between the initial threshold of each key node and the historical average is calculated. If the deviation is within 5%, the initial threshold is retained. If the deviation exceeds 5%, the initial threshold is fine-tuned in conjunction with historical data. For example, the initial planned time of 90 days for the completion of core module coding deviates by 2.2% from the historical average of 88 days, which is within the allowable range, so the 90-day time threshold is retained. The initial threshold of 90% for the approval rate of requirement documents deviates by 1.1% from the historical average of 91%, so the initial threshold is retained. The initial threshold of 99.8% for the integrity of historical data migration deviates by 0.1% from the historical average of 99.7%, so the initial threshold is retained. This training process ensures that the threshold settings conform to the actual project execution rules and avoids unreasonable thresholds caused by deviating from historical experience.

[0034] After training, a project plan structure model is established. The optimized key node definitions, time thresholds, and quality thresholds are categorized and organized according to project stages. A tree structure is used to construct the model framework, with the root node set as the manufacturing group ERP system upgrade project. The first-level sub-nodes correspond to the five project stages: requirements research, system development, module testing, data migration, and trial operation. Under each first-level sub-node, second-level sub-nodes are created that are consistent with the key node definitions of that stage. For example, under the first-level sub-node of the requirements research stage, a second-level sub-node for requirements document review completion is created; under the first-level sub-node of the system development stage, a second-level sub-node for core module coding completion is created. The optimized time and quality threshold data are associated one by one with the corresponding second-level sub-nodes. For example, the second-level sub-node for requirements document review completion is associated with the threshold information of the 30th day of the planned time and a review pass rate of not less than 90%, and the second-level sub-node for core module coding completion is associated with the threshold information of the 90th day of the planned time and a coding pass rate of not less than 98%.

[0035] Finally, the constructed tree structure model is stored in the model database of the project supervision management system. At the same time, the model call interface is configured. The interface supports quick querying of corresponding threshold data by project stage name and key node name, ensuring that the optimized real-time data in subsequent steps can be matched and called by task identifier to call the node and threshold information in the model, so as to achieve seamless connection between the model and the data processing flow.

[0036] Step 302: Based on the project plan structure model, match and map the optimized real-time data with key nodes according to task identifiers to establish the correspondence between data and plan nodes. Specifically, this includes: extracting the task identifier of each data point from the optimized real-time data. This task identifier is generated by the data source system. For example, the identifier of the requirement document review task in the client's OA system is OA-ERP-REQ-001, the identifier of the core module coding task in the vendor's Jira system is Jira-ERP-DEV-003, the identifier of the test case execution task in the third-party TestRail system is TestRail-ERP-TST-005, the identifier of the historical data migration task in the data migration tool is Mig-ERP-DAT-002, and the identifier of the trial operation monitoring task in the system operation and maintenance platform is Ops-ERP-RUN-001; constructing a source point set from all extracted task identifiers, with each task identifier as an independent element in the source point set, and each element having attributes such as the data source system type and task creation time.

[0037] The key node information in the project plan structure model is called up to construct a target point set, which includes key nodes such as the completion of requirement document review in the requirement research phase, the completion of core module coding in the system development phase, the achievement of test case pass rate in the module testing phase, the completion of historical data migration in the data migration phase, and the stable operation of the system for 72 hours in the online trial operation phase. Each key node is an independent element in the target point set, and each element is accompanied by attributes such as the project phase to which it belongs and the associated threshold type.

[0038] A point-set mapping algorithm is used to process the source and target point sets. First, feature information is extracted from each task identifier element in the source point set, including key fields in the identifier string such as REQDEVTSTDATRUN, and the data source system prefix such as OAJiraTestRailMigOps. Feature information is extracted from each key node element in the target point set, including core descriptive terms in the node name such as REQ for requirement document review, DEV for core module coding, TST for test cases, DAT for data migration, and RUN for stable operation, as well as the stage identifier of the node such as the requirement survey system development module test data migration online trial operation.

[0039] The feature similarity between source point set elements and target point set elements is calculated using a point set mapping algorithm. For each source point set element, its key fields, system prefixes, and core descriptive terms and stage identifiers are compared with those of the target point set element. For example, the key field REQ of the source point set element OA-ERP-REQ-001 matches the core descriptive term REQ of the target point set element "requirement document review completed". The system prefix OA is consistent with the commonly used associated system OA of the requirement survey stage to which the target point set element belongs. The similarity between the two is calculated. A similarity threshold is set, and only source point set and target point set elements with similarity exceeding the preset threshold are paired and combined to establish a mapping rule between task identifiers and key nodes.

[0040] Based on the above mapping rules, each piece of optimized real-time data is matched. For example, the source point set element OA-ERP-REQ-001 is associated with the target point set element "requirement document review completed"; the source point set element Jira-ERP-DEV-003 is associated with the target point set element "core module coding completed"; the source point set element TestRail-ERP-TST-005 is associated with the target point set element "test case pass rate met"; the source point set element Mig-ERP-DAT-002 is associated with the target point set element "historical data migration completed"; and the source point set element Ops-ERP-RUN-001 is associated with the target point set element "system stable operation for 72 hours". After all data matching is completed, a one-to-one correspondence table between each piece of real-time data and the corresponding key node is formed. The table includes the source point set element identifier, the target point set node name, the feature similarity value, and the data source system information.

[0041] Step 303: Based on the aforementioned correspondence, perform time-dimensional analysis on the progress data of each key node, calculate the difference between the actual completion time and the planned time, and obtain the time deviation data. Specifically, this includes: retrieving the optimized real-time data associated with each key node from the correspondence table established in Step 302, and extracting the actual completion time from the data. For example, in the real-time data corresponding to the requirement document review completion node, the actual completion time is the 32nd day after project start; in the real-time data corresponding to the core module coding completion node, the actual completion time is the 88th day after project start; in the real-time data corresponding to the test case pass rate achievement node, the actual completion time is the 153rd day after project start; in the real-time data corresponding to the historical data migration completion node, the actual completion time is the 181st day after project start; and in the real-time data corresponding to the system stable operation for 72 hours node, the actual completion time is the 208th day after project start. Simultaneously, retrieve the data for each key node from the project plan structure model. The planned time thresholds for each key node are as follows: the planned time for completing the requirements document review is day 30; the planned time for completing the core module coding is day 90; the planned time for achieving the test case pass rate target is day 150; the planned time for completing the historical data migration is day 180; and the planned time for the system to run stably for 72 hours is day 210. For each key node, the difference between the actual completion time and the planned time is calculated. For example, the time difference for completing the requirements document review node is 32-30=2 days; the time difference for completing the core module coding node is 88-90=-2 days; the time difference for achieving the test case pass rate target is 153-150=3 days; the time difference for completing the historical data migration node is 181-180=1 day; and the time difference for the system to run stably for 72 hours is 208-210=-2 days. The node name, planned time, actual completion time, and time difference for each key node are compiled into a table to form the time deviation data.

[0042] Step 304: Based on the time deviation data, perform numerical dimension analysis on the quality indicators of each key node, calculate the percentage difference between the actual and planned numerical indicators, and obtain the numerical difference rate data. Specifically, based on the time deviation data obtained in step 303, for each key node, extract the corresponding actual numerical indicators from the optimized real-time data. For example, the actual numerical indicator for the requirement document review completion node is a requirement document review pass rate of 92%; the actual numerical indicator for the core module coding completion node is a core module coding pass rate of 97.5%; the actual numerical indicator for the test case pass rate meeting node is a test case pass rate of 93%; the actual numerical indicator for the historical data migration completion node is historical data migration integrity of 99.7%; and the actual numerical indicator for the system trial operation stability node is system trial operation failure rate of 0.12%. Retrieve the quality thresholds corresponding to each key node from the project plan structure model, namely, a requirement document review pass rate of 90%, a core module coding pass rate of 98%, a test case pass rate of 95%, and a historical data migration integrity of 99.7%. The system achieved a pass rate of 99.8% and a failure rate of 0.1% during trial operation. For each key node, the percentage difference between the actual and planned numerical indicators was calculated as (actual numerical indicator - planned numerical indicator) ÷ planned numerical indicator × 100%. For example, the percentage difference for the requirement document review completion node was (92% - 90%) ÷ 90% × 100% ≈ 2.22%, and the percentage difference for the core module coding completion node was (97.5% - 98%) ÷ 98% × 100% ≈ -0.51%. The test cases passed... The percentage difference for nodes that meet the pass rate target is (93%-95%)÷95%×100%≈-2.11%, the percentage difference for nodes that have completed historical data migration is (99.7%-99.8%)÷99.8%×100%≈-0.10%, and the percentage difference for nodes that are stable during system trial operation is (0.12%-0.1%)÷0.1%×100%=20%. The node name, planned numerical indicators, actual numerical indicators, and percentage difference for each key node are compiled into a table to form the numerical difference rate data.

[0043] Step 305: Based on the time deviation data and numerical difference rate data, integrate all deviation calculation results to form a comparison result. Specifically, this includes: associating and integrating the time deviation data and numerical difference rate data according to key nodes. Each key node corresponds to an integrated record, and the record content includes the node name, planned time, actual completion time, time deviation, planned numerical indicator, actual numerical indicator, and numerical difference rate. Each integrated record is marked with a status. If the time deviation is positive and its absolute value is greater than 3 days, or the numerical difference rate is negative and its absolute value is greater than 5%, it is marked as a status requiring attention. If the time deviation is positive and its absolute value is between 1 and 3 days, or the numerical difference rate is negative and its absolute value is between 2% and 5%, it is marked as... The status is to be observed. If the absolute value of the time deviation is less than 1 day and the absolute value of the numerical difference rate is less than 2%, it is marked as normal. For example, if the time deviation of the requirement document review completion node is 2 days and the numerical difference rate is 2.22%, it is marked as to be observed; if the time deviation of the core module coding completion node is -2 days and the numerical difference rate is -0.51%, it is marked as normal; if the time deviation of the test case pass rate meeting the standard is 3 days and the numerical difference rate is -2.11%, it is marked as needing attention; if the time deviation of the historical data migration completion node is 1 day and the numerical difference rate is -0.10%, it is marked as normal; if the time deviation of the system trial operation stabilization node is -2 days and the numerical difference rate is 20%, it is marked as needing attention. All the marked integrated records are sorted by project stage to form a complete comparison result.

[0044] In a preferred embodiment of the present invention, step 400 above involves real-time early warning judgment based on the comparison results. When a schedule deviation or key indicator failure is identified, an early warning event is automatically generated, and its transmission impact on the project's critical path is analyzed through a pre-configured rule engine to obtain early warning information, including: Step 401: Based on the comparison results, match and map the time deviation data with the preset time tolerance threshold to obtain a set of deviation status identifiers. Simultaneously, compare and analyze the numerical difference rate data with the corresponding quality pass standards to obtain comparison analysis results. Specifically, this includes: Based on the comparison results, firstly, retrieve the preset time tolerance threshold for the manufacturing group ERP system upgrade project. This threshold is set according to the characteristics of different project stages. Specifically, the time tolerance threshold for the demand survey stage and system development stage is ±2 days, the time tolerance threshold for the module testing stage and data migration stage is ±3 days, and the time tolerance threshold for the online trial operation stage is ±1 day. Then, match and map the time deviation data of each key node in the comparison results with the corresponding time tolerance threshold. The matching and mapping rule is that if the time deviation is within the tolerance range, the node is marked with a normal status identifier; if the time deviation exceeds the tolerance limit, the node is marked with a positive out-of-tolerance status identifier. If the deviation is below the lower tolerance limit, the node is marked with a negative out-of-tolerance status. For example, the time deviation for the requirement document review completion node is 2 days, corresponding to the tolerance threshold of ±2 days in the requirement research phase, so the node is marked with a normal status. The time deviation for the test case pass rate achievement node is 3 days, corresponding to the tolerance threshold of ±3 days in the module testing phase, so the node is marked with a normal status. The time deviation for the historical data migration completion node is 1 day, corresponding to the tolerance threshold of ±3 days in the data migration phase, so the node is marked with a normal status. The time deviation for the core module coding completion node is -2 days, corresponding to the tolerance threshold of ±2 days in the system development phase, so the node is marked with a normal status. The time deviation for the system stable operation for 72 hours is -2 days, corresponding to the tolerance threshold of ±1 day in the online trial operation phase, so the node is marked with a negative out-of-tolerance status. The status indicators of all key nodes are integrated together to form a deviation status indicator set.

[0045] Simultaneously, the quality qualification standards corresponding to each key node are retrieved. These quality qualification standards are consistent with the quality thresholds extracted from the project plan in step 301, specifically including a requirement document review pass rate of no less than 90%, a core module coding pass rate of no less than 98%, a test case pass rate of no less than 95%, historical data migration integrity of no less than 99.8%, and a system trial operation failure rate of no more than 0.1%. Then, the numerical difference rate data of each key node in the comparison results are compared and analyzed with the corresponding node's quality qualification standards. The rule for comparison and analysis is that if the numerical difference rate is positive or zero, and the actual numerical indicator of the node meets the requirements of the quality qualification standard, then the node is judged to be qualified; if the numerical difference rate is negative, and the actual numerical indicator of the node does not meet the requirements of the quality qualification standard, then the node is judged to be unqualified. For example, the numerical difference rate of the requirement document review pass rate is 2.22%, and the actual pass rate is... The pass rate was 92%, meeting the quality standard of no less than 90%, so this node was deemed compliant. The pass rate of the core module coding had a numerical difference rate of -0.51%, with an actual pass rate of 97.5%, failing to meet the quality standard of no less than 98%, so this node was deemed non-compliant. The pass rate of the test cases had a numerical difference rate of -2.11%, with an actual pass rate of 93%, failing to meet the quality standard of no less than 95%, so this node was deemed non-compliant. The completeness of historical data migration had a numerical difference rate of -0.10%, with an actual completeness of 99.7%, failing to meet the quality standard of no less than 99.8%, so this node was deemed non-compliant. The failure rate of the system trial operation had a numerical difference rate of 20%, with an actual failure rate of 0.12%, failing to meet the quality standard of no more than 0.1%, so this node was deemed non-compliant. The results of all the key nodes were compiled together to form a comparative analysis.

[0046] Step 402: Based on the set of deviation status identifiers and the comparison analysis results, a warning condition judgment is performed. When the time deviation exceeds the time tolerance threshold or the numerical difference rate is lower than the quality pass standard, a warning event is automatically generated. Specifically, this includes: pre-setting warning conditions, i.e., when the status identifier of a key node is positive or negative out-of-tolerance, or the comparison analysis result is substandard, a warning event is triggered; based on the set of deviation status identifiers and the comparison analysis results obtained in step 401, the system automatically traverses the status information of all key nodes and checks whether the warning conditions are met one by one; for example, if the comparison analysis result of the core module coding completion node is substandard, the warning conditions are met, and the system automatically generates a warning event for that node. The event content includes the node name, warning type, actual numerical index, quality pass standard, and numerical difference rate. Among them, the node name is "core module coding completed," the warning type is "quality substandard," the actual numerical index is 97.5%, the quality pass standard is 98%, and the numerical difference rate is -0.5%. 1%; The test case pass rate meets the standard, but the comparison analysis result is substandard, meeting the warning conditions. An warning event is generated, including the node name, warning type, actual numerical indicator, quality pass standard, and numerical difference rate. Among them, the node name is "Test Case Pass Rate Meets Standard," the warning type is "Quality Substandard," the actual numerical indicator is 93%, the quality pass standard is 95%, and the numerical difference rate is -2.11%. The system has been running stably for 72 hours, and the node status is negatively out of tolerance. The comparison analysis result is substandard, generating a warning event including the node name, warning type, actual completion time, planned time, time deviation, actual numerical indicator, quality pass standard, and numerical difference rate. Among them, the node name is "System Has Been Running Stably for 72 Hours," the warning type is "Time Deviation and Quality Substandard," the actual completion time is the 208th day after project start, the planned time is the 210th day, the time deviation is -2 days, the actual numerical indicator is 0.12%, the quality pass standard is 0.1%, and the numerical difference rate is 20%.

[0047] Step 403: Based on the aforementioned warning event, critical path impact analysis is performed using a pre-configured rule engine to obtain the critical path impact analysis results. Specifically, for the ERP system upgrade project of the manufacturing group, the key nodes of the entire project process are first identified, including the completion of requirement document review, completion of core module coding, achievement of test case pass rate, completion of historical data migration, and 72 hours of stable system operation. Then, through the project plan specification and multi-party collaborative records, the pre-dependencies and post-influence relationships of each node are determined to form a dependency chain. That is, the completion of requirement document review is the only pre-dependency node of core module coding completion, the completion of core module coding is the only pre-dependency node of test case pass rate achievement, the achievement of test case pass rate achievement is the only pre-dependency node of historical data migration completion, and the completion of historical data migration is the only pre-dependency node of 72 hours of stable system operation. This dependency chain is then entered into the node relationship database of the rule engine.

[0048] In the rule engine, the delay propagation logic of the preceding node is set to ensure that the actual delay of the preceding node is directly equal to the startup delay of the following node, and this delay will be propagated sequentially along the dependency chain. For example, if the preceding node is delayed by N days, the startup time of the following node will be synchronously delayed by N days, and the actual completion time of the following node = the original planned completion time + N days. If the following node has its own delay, the total delay = the delay propagated by the preceding node + its own delay. This rule is bound to the node dependency relationship for storage.

[0049] Configure the basic parameters required for impact analysis in the rules engine, including the original planned start time, original planned completion time, and node weight of each key node. For example, set the weight to the highest if the system has been running stably for 72 hours. At the same time, set the delay threshold judgment standard. That is, when the delay time of a node exceeds 50% of the time tolerance threshold of the corresponding stage, it should be marked and its transmission impact should be analyzed first to ensure that subsequent analysis can quickly locate key impact nodes.

[0050] Based on the warning events, the rule engine first automatically retrieves the pre-configured node relationship database, delay propagation rules, and impact analysis parameters to conduct impact analysis on each warning event. For example, for a warning event where the test case pass rate meets the standard, the rule engine first extracts the actual delay duration of the node from the warning event. Assuming that subsequent investigation determines it to be 3 days, it then locates its downstream node as the historical data migration completion through the node relationship database. According to the pre-configured delay propagation rules, it is determined that the 3-day delay of the test case pass rate meeting the standard will directly cause the start time of the historical data migration completion to be delayed by 3 days. Combining the original planned start time of the historical data migration completion (assumed to be day 150) and the original planned completion time (day 180), it is calculated that the actual start time of the node becomes day 153 and the actual completion time becomes day 183.

[0051] The rule engine then continues its analysis along the dependency chain. By using the node relationship database, it determines that the subsequent node after the historical data migration is completed is the one where the system has been running stably for 72 hours. Applying the delay propagation rule again, it determines that the 3-day delay in the completion of the historical data migration will delay the start time of the system's stable 72-hour operation from the originally planned day 181 to day 184, and the actual completion time from the originally planned day 210 to day 213. At the same time, the rule engine checks whether the 3-day delay of the node with the test case pass rate exceeds 50% of the time tolerance threshold (±3 days) of its corresponding stage (module testing stage), which is 1.5 days, according to the pre-configured delay threshold judgment standard. Since 3 days is greater than 1.5 days, the node is marked as a key impact source.

[0052] Finally, the key information from the above analysis process is compiled into the critical path impact analysis results, which include the names of affected nodes, the original planned start and finish times of each node, the expected delayed start and finish times, the delay propagation path, and the markers of key impact sources. Among them, the names of affected nodes are those where historical data migration is completed and the system has been running stably for 72 hours. The delay propagation path is that the delay of the node that meets the test case pass rate standard is propagated sequentially to the node where historical data migration is completed, and then to the node where the system has been running stably for 72 hours. The markers of key impact sources are the nodes where the test case pass rate standard is met.

[0053] Step 404: Based on the critical path impact analysis results, calculate the impact of the warning event on the overall project schedule and obtain the warning information. Specifically, this includes: Based on the critical path impact analysis results, the system first identifies the total length of the critical path affected by the warning event. The total critical path duration of the manufacturing group ERP system upgrade project is 210 days, and the system will be completed within 72 hours from project initiation to stable operation. Then, the system calculates the total delay time of the critical path caused by the warning event. For example, if the test case pass rate target node is delayed by 3 days, it will cause the subsequent two critical nodes to be delayed by 3 days each. The total delay time of the critical path is consistent with the maximum delay time of the preceding nodes, i.e., 3 days. At the same time, the system determines the degree of impact in conjunction with the remaining project cycle. If the current project has been running for 150 days and there is a remaining 60-day cycle, a 3-day delay accounts for 5% of the remaining cycle, and the impact is judged to be moderate. If the delay reaches 5 days, accounting for 8.3% of the remaining cycle, the impact is judged to be significant. Finally, the warning event information, critical path impact analysis results, total delay duration, and impact degree are integrated to form a warning information. The content includes the warning event number, associated critical nodes, warning type (time deviation / quality non-compliance), list of critical path affected nodes, total delay duration, impact degree level, and estimated final project completion time (originally planned for day 210, delayed to day 213), ensuring that the warning information can fully reflect the impact of the warning event on the overall project progress.

[0054] In a preferred embodiment of the present invention, step 500 above involves pushing the warning information to the systems of Party A and Party B through a message middleware and collecting feedback status to complete closed-loop collaborative processing. Simultaneously, the processing results are updated to the central database, including: Step 501: Based on the aforementioned warning information, format the message using a message middleware to obtain a standardized warning message. Specifically, this includes: Based on the warning information, which contains detailed information about each warning node in the manufacturing group's ERP system upgrade project, such as the warning type (quality substandard, actual numerical indicator 97.5%, quality pass standard 98%) for the core module coding completion node; and the warning type (time deviation + quality substandard) for the system stable operation 72-hour node, with actual completion time being the 208th day after project start and planned time being the 210th day. The message middleware's formatting function is invoked to predefine the structure of the standardized warning message, which includes basic project information fields such as project identifier, project name, and project... In this phase, the warning event fields include Warning Event ID, Node Name, and Warning Type; Deviation Details fields include Actual Value, Planned Value, and Deviation Amount; and Recipient Identifier fields include Receiving System Type and Receiving Interface Address. Each data point in the warning information is filled in according to the above structure. The Project Identifier is the unique number of the manufacturing group's ERP system upgrade project. The Receiving System Type is labeled as both the client's management system and the vendor's development system. The Receiving Interface Address is the progress warning receiving interface address for the client's OA system and the development task warning interface address for the vendor's Jira system. After data filling is complete, the message middleware automatically performs format validation on the message to ensure no fields are missing and the data format is consistent, ultimately generating a standardized warning message.

[0055] Step 502: Based on the standardized early warning message, push and transmit the data to the designated interfaces of the client's management system and the vendor's development system through the message middleware, and obtain the push and transmission results. Specifically, based on the standardized early warning message, the message middleware first collects information on available transmission paths for communication with the client's management system and the vendor's development system, including parameters such as the historical transmission time node load rate and data packet loss rate of each path; applies a multi-objective shortest path algorithm to perform weighted calculations on these parameters, where the transmission time weight is set to 0.4, the node load rate weight is set to 0.3, and the data packet loss rate weight is set to 0.3. The algorithm selects the transmission path with the best overall performance, that is, for the client's management system, select the path with the shortest transmission time and a load rate of less than 30%, and for the vendor's development system, select the path with a packet loss rate of less than 1% and the highest transmission stability score.

[0056] After determining the optimal path, the message middleware establishes communication connections with the client's management system and the vendor's development system based on this path. During the connection process, a preset encryption protocol is used to ensure data transmission security. For the client's management system, the message middleware pushes standardized early warning messages through the selected path to its preset progress early warning receiving interface. This interface is a dedicated interface in the client's OA system for receiving supervisory early warning information. For the vendor's development system, standardized early warning messages are pushed through the corresponding optimal path to its preset development task early warning interface. This interface is associated with the task management module in the vendor's Jira system.

[0057] During the push process, the message middleware monitors the status parameters of the transmission path in real time. If the transmission time exceeds the expected threshold or the node load rate suddenly increases, the path recalculation mechanism is automatically triggered, the multi-target shortest path algorithm is called to update the optimal path and switch the transmission channel. If a transmission timeout or transmission failure occurs, the retransmission mechanism is automatically started, with the retransmission interval set to 30 seconds and a maximum of 3 retransmissions.

[0058] After transmission is completed, the message middleware records the push transmission result of each message. The result includes the name of the receiving system of the push object, the transmission path identifier, the transmission status (success or failure), the transmission time, the number of retransmissions, and the number of path switches, forming a push transmission result list.

[0059] Step 503: Based on the push transmission results, collect the feedback status information returned by the systems of Party A and Party B, and parse and process the feedback status information to obtain a status response report. Specifically, this includes: based on the push transmission results, the message middleware continuously monitors the feedback ports of Party A's management system and Party B's development system, with a monitoring interval of 1 minute; when Party A's management system receives an early warning message, its OA system automatically generates feedback status information; this feedback status information includes the feedback system identifier as Party A's management system, the early warning event ID, the feedback status as viewed or being processed, the feedback time, and the person responsible for processing; when Party B's development system receives an early warning message, its Jira system automatically associates the corresponding development task and generates feedback status information; this feedback status information includes the feedback system identifier as Party B's development system, the early warning event ID, the feedback status as confirmed or a solution has been formulated, the feedback time, the person responsible for processing, and contact information.

[0060] After receiving the aforementioned feedback status information, the message middleware extracts the feedback timestamp and status code from each message as data points. The feedback timestamp is converted into a continuous numerical status code and quantified according to "viewed" (0), "confirmed" (1), "processing" (2), and "plan formulation" (3). The convex hull algorithm is applied to calculate these data points to determine the minimum convex polygon boundary containing all valid data points. Data points within the boundary are considered normal feedback data, while data points outside the boundary are marked as abnormal feedback data. The normal feedback data within the convex hull boundary is parsed, key fields are extracted, redundant information is removed, and the feedback from Party A and Party B for the same warning event is associated. For abnormal feedback data outside the convex hull boundary, the abnormality type is marked, such as feedback timestamp exceeding a reasonable range or invalid status code, and recorded separately.

[0061] The normal feedback information after being categorized and analyzed according to the warning event ID is combined with the marking results of abnormal feedback data to form a status response report. Each record in the report includes the warning event ID, node name, feedback status of Party A, responsible person of Party A, feedback status of Party B, responsible person of Party B, feedback time, and abnormality marker.

[0062] Step 504: Based on the status response report, perform correlation analysis with the early warning information to complete the closed-loop collaborative processing of the early warning event, obtain the processing result, and update the processing result to the central database. Specifically, this includes: based on the status response report, extracting the core features of the feedback status information in the report, including the expected response time of the person responsible for the feedback status processing measures; simultaneously extracting the core features of the early warning information obtained in step 404, including the early warning type and the importance and processing requirements of the early warning level nodes; mapping the two types of features into multi-dimensional space vectors, where the feedback status is quantified as viewed (0), confirmed (1), processing (2), and formulating a plan (3), and the early warning type is quantified as time deviation (0), quality failure (1), and compound deviation (2); other features are converted into values ​​in the range of 0 to 1 according to preset rules.

[0063] A spatial similarity matching algorithm is applied to calculate the cosine similarity between two types of vectors, with a similarity threshold of 0.7. If the result is greater than or equal to 0.7, the feedback and early warning requirements are considered to match; otherwise, they are considered not to match. For each early warning event, the collaborative processing status is confirmed based on the similarity result. For example, if the early warning information feature vector of the core module coding completion node has a similarity of 0.85 with the feedback feature vector of the solution formulated by the contractor, it is considered a match and the processing is progressing normally. If the similarity between the feedback feature vector and the early warning information feature vector of a certain early warning event is 0.6, it is marked as an insufficient match and a suggestion to supplement the feedback is made. All early warnings are integrated. The similarity matching results and processing details of the events are used to clarify the current processing stage, processing measures, responsible persons, expected completion time, and matching status of each early warning event, thus forming a processing result. The write interface of the central database is called to update the early warning event processing table in the database with the processing result. This table contains fields such as early warning event ID, processing result, processing stage, processing measures, responsible persons, expected completion time, matching degree score, and update time. After the update is completed, the central database automatically generates an update log, recording the updated content and update time of the processing result and the matching degree analysis result, ensuring data traceability and completing the closed-loop collaborative processing of early warning events.

[0064] In a preferred embodiment of the present invention, step 600, based on the central database and combined with historical project data, performs pattern recognition analysis to extract schedule deviation patterns, and dynamically optimizes the planned node thresholds and buffer times for subsequent stages to obtain an optimization strategy, including: Step 601: Based on the central database and combined with the historical project database, extract the current project progress data and historical project progress data, and perform feature engineering processing on the data to obtain a standardized feature dataset. Specifically, this includes: extracting the progress data of the current manufacturing group's ERP system upgrade project based on the central database, including information such as the actual completion time, planned time, time deviation, actual quality indicators, and numerical difference rate of each stage's key nodes; and simultaneously connecting to the historical project database to extract the progress data of similar manufacturing ERP system upgrade projects within the past three years, covering historical project data of different enterprise sizes, project cycles, and participating teams, such as the progress deviation records of an ERP upgrade project for an automobile manufacturing company and an ERP upgrade project for a machinery manufacturing company.

[0065] Feature engineering was performed on the extracted current and historical project progress data. First, the data format was standardized by converting the time data to a uniform format based on the Xth day after the project started, and the quality indicators (such as pass rate and qualification rate) were uniformly quantified to values ​​from 0 to 100. Second, missing data was processed by filling in the missing quality indicator data with the mean of similar projects in the same stage, and by supplementing the missing time deviation data with linear interpolation of the deviation between adjacent nodes. Finally, data standardization was performed by converting the time deviation to a value between 0 and 1 according to the proportion of the project cycle, and scaling the quality indicator difference rate to the range of -1 to 1 according to the calculation result of (actual value - standard value) / standard value. The final result is a standardized feature dataset containing the characteristics of current and historical projects.

[0066] Step 602: Based on the standardized feature dataset, a clustering analysis method is used for pattern recognition to identify typical patterns of schedule deviations and obtain a set of schedule deviation patterns. Specifically, this includes: Based on the standardized feature dataset, a K-means clustering analysis method is used for pattern recognition. First, the number of clusters is determined by the elbow rule, and the number of clusters is initially set to 3 based on the types of schedule deviations in historical projects. The standardized values ​​of time deviation, standardized values ​​of quality indicator difference rates, frequency of requirement changes, and development team size are used as core clustering features. These features are input into the K-means algorithm for iterative calculation until the cluster centers are stable. After clustering, feature analysis was performed on each clustering result to identify typical patterns of schedule deviations: Pattern 1 is time deviation-dominated, in which more than 80% of the deviations in the data come from time deviations exceeding the threshold, while the difference rate of quality indicators is mostly within the acceptable range, which is common in the early requirements survey stage of the project; Pattern 2 is quality non-compliance-dominated, in which more than 75% of the deviations are manifested by quality indicators failing to meet the standards, while the time deviation is relatively small, which mainly occurs in the system development and module testing stages; Pattern 3 is composite deviation, in which time deviation and quality deviation coexist, and the degree of deviation is relatively high, which mostly occurs in the data migration and online trial operation stages; The three typical patterns and their corresponding clustered data samples were organized to form a schedule deviation pattern set.

[0067] Step 603: Based on the set of schedule deviation patterns, perform statistical analysis on the deviation characteristics of each pattern, extract the distribution pattern and influencing factors of schedule deviation, and obtain a schedule deviation pattern analysis report. Specifically, this includes: based on the set of schedule deviation patterns, performing statistical analysis on the deviation characteristics of each pattern: for the time deviation-dominant pattern, calculating the mean, median, and standard deviation of the time deviation for all samples under this pattern, finding that the time deviation under this pattern is mostly concentrated in the 2-5 day range, and that for every increase in the frequency of requirement changes, the time deviation increases by an average of 1.2 days; for the quality non-compliance-dominant pattern, statistically analyzing the distribution range of the quality indicator difference rate under this pattern, and... Currently, 90% of the sample variance rate is concentrated in the range of -5% to -2%, and when the proportion of developers with ERP project experience is less than 60%, the probability of this pattern increases by 40%. For the compound deviation pattern, the analysis shows that this pattern mostly occurs in the second half of the project cycle, and when the data migration volume exceeds 500GB, the probability of compound deviation increases by 35%. At the same time, the influencing factors of each pattern are extracted, including the frequency of requirement changes, the experience level of the development team, the data scale, and the investment of testing resources. The statistically obtained deviation distribution range, the correlation of influencing factors, and the probability of each pattern occurring in different project stages are compiled into a schedule deviation pattern analysis report.

[0068] Step 604: Based on the progress deviation pattern analysis report, predictive analysis methods are used to predict the progress trend of subsequent stages to obtain prediction results; the planned node thresholds and buffer time settings are dynamically adjusted according to the prediction results to obtain optimization strategies. Specifically, based on the progress deviation pattern analysis report, time series analysis methods are used to predict the progress trend of subsequent stages of the current ERP system upgrade project. Based on the progress change patterns of the same model in historical projects, combined with the deviation data of the completed stages of the current project, it is predicted that the subsequent module testing stage may have a quality non-compliance-dominant deviation, and the predicted time deviation is about 1 day and the quality difference rate is about negative 3%.

[0069] Based on the forecast results, dynamically adjust the planned node thresholds and buffer time settings for subsequent stages: adjust the time tolerance threshold for the module testing stage from ±3 days to ±4 days, adding 1 day of buffer time to cope with the predicted time deviation; slightly adjust the quality pass standard for the module testing stage from a test case pass rate of no less than 95% to no less than 94%, and require the testing team to add 2 personnel with ERP testing experience; optimize the planned node thresholds for the data migration stage, keeping the historical data migration integrity threshold unchanged from 99.8%, but adding 1 day of data verification buffer time; compile the adjusted planned node thresholds, buffer time settings, and corresponding adjustment basis to form a progress control optimization strategy for subsequent stages.

[0070] In a preferred embodiment of the present invention, step 700, based on an optimization strategy, involves constructing a configurable progress control template library to quickly adapt to and generate corresponding progress control nodes and early warning thresholds in new types of projects, including: Step 701: Based on the optimization strategy, analyze the planned node thresholds, buffer time parameters, and early warning condition configurations contained therein to obtain a standardized configuration parameter set. Specifically, this includes: Based on the optimization strategy, first extracting the planned node thresholds for each stage of the manufacturing group ERP system upgrade project. For example, the planned node threshold for the demand survey stage is adjusted from 30 days to 28 days, the system development stage from 60 days to 58 days, the module testing stage from 30 days to 32 days, the data migration stage from 15 days to 16 days, and the online trial operation stage from 15 days to 14 days; then extracting the buffer time parameters corresponding to each stage, increasing the buffer time for the demand survey stage from 2 days to... The system development phase has been extended from 3 days to 4 days, the module testing phase from 2 days to 3 days, and the data migration phase from 1 day to 2 days. The online trial operation phase remains unchanged at 1 day. At the same time, the optimized early warning condition configuration has been extracted. For example, a time early warning is triggered when the time deviation exceeds ±1.5 days, a quality early warning is triggered when the numerical difference rate is less than -1.2%, and a priority early warning is triggered when there is a compound deviation (time deviation and quality failure). These planned node thresholds, buffer time parameters, and early warning condition configurations are classified and organized according to the structure of project phase-parameter type-numerical range-configuration description. Duplicate or conflicting parameter content is removed, and parameter formats and units are unified to form a standardized set of configuration parameters.

[0071] Step 702: Based on the standardized configuration parameter set, construct a configurable progress control template library according to project type. The progress control template library includes node definitions, threshold settings, and early warning rules. Specifically, based on the standardized configuration parameter set, first divide the project type into three categories: ERP system upgrade projects, data center construction projects, and supply chain management system deployment projects. For each type of project, construct a corresponding configurable progress control template. The ERP system upgrade project template includes node definitions for five stages: requirements gathering, system development, module testing, data migration, and trial operation. Each node is associated with a corresponding node. The project template includes planned node thresholds, buffer time parameters, and rules for time-based, quality-based, and composite early warnings. The data center construction project template includes node definitions for five stages: data center renovation, equipment procurement, equipment deployment, system debugging, and data migration, along with corresponding thresholds and early warning rules. The supply chain management system deployment project template includes node definitions and related parameters for five stages: requirements analysis, system customization development, interface integration, user training, and online operation. These three types of project templates are integrated to form a progress control template library. Each template in the library includes a node definition module, a threshold setting module, and an early warning rule module, and the parameters of each module can be adjusted according to the needs of new projects.

[0072] Step 703: Based on the progress control template library, when applied to a new type of project, extract the characteristic parameters of the new project; based on the characteristic parameters of the new project, match and map them with the baseline template in the progress control template library to determine the final matching baseline template. Specifically, this includes: based on the progress control template library, when facing a new type of project, such as the deployment project of a supply chain management system for an automobile manufacturing enterprise, first extract the characteristic parameters of the new project, including the project cycle of 8 months, involving five stages: requirements analysis, system customization and development, interface integration, user training, and online operation; the participating parties include the client's information department, the vendor's development team, and a third-party interface testing organization; and the management systems used by each party. The system comprises the client's OA system, the vendor's Jira system, and the third-party TestRail system. Key deliverables include a requirements specification, system source code, interface test reports, and user training manuals. These characteristic parameters were then matched against a baseline template for a supply chain management system deployment project in the template library. The comparison examined whether the number of phases was consistent, whether the participating management system types were compatible, and whether the nodes corresponding to the key deliverables matched. The comparison showed that the new project's phase divisions and management system types overlapped with the baseline template by over 90%, and the nodes corresponding to the key deliverables were a perfect match. Therefore, this baseline template for the supply chain management system deployment project was determined to be the final matching baseline template.

[0073] Step 704: Based on the final matched baseline template, generate customized progress control nodes through parameter adaptation and adjustment. Specifically, this includes: based on the supply chain management system deployment project baseline template determined in Step 703, first retrieve the planned node thresholds and buffer time parameters for each stage in the template. For example, in the baseline template, the planned node threshold for the demand analysis stage is 30 days with a buffer time of 3 days; the planned node threshold for the system customization development stage is 90 days with a buffer time of 10 days; the planned node threshold for the interface integration stage is 45 days with a buffer time of 5 days; the planned node threshold for the user training stage is 30 days with a buffer time of 3 days; and the planned node threshold for the online operation stage is 15 days with a buffer time of 2 days. Combined with the actual needs of the new project, the client proposes... Custom system development requires an additional 10 days to adapt to the company's specific supply chain processes, and interface integration requires an additional 5 days to connect with 3 external supplier systems. Therefore, the planned node threshold for the custom system development phase is adjusted to 100 days with a buffer period of 10 days, and the planned node threshold for the interface integration phase is adjusted to 50 days with a buffer period of 6 days. At the same time, based on the specific requirements of the key deliverables of the new project, the nodes under the custom system development phase are refined into three sub-nodes: completion of core function development, completion of special process adaptation, and completion of system integration testing. The nodes under the interface integration phase are refined into three sub-nodes: completion of interface integration with supplier A, completion of interface integration with supplier B, and completion of interface integration with supplier C. Finally, customized progress control nodes are generated.

[0074] Step 705: Based on the customized progress control nodes, generate corresponding early warning thresholds using the early warning rule engine. Specifically, this includes: based on the customized progress control nodes, calling the pre-configured early warning rule engine; first, determining the early warning logic corresponding to each node in the rule engine, i.e., setting the early warning deviation range for the time dimension as 5% of the planned node threshold, and setting the numerical difference rate early warning range for the quality dimension as 1% of the key indicator qualification standard; for special process adaptation completion nodes, whose planned completion time is the 50th day of the system customization development phase, with a total planned time of 100 days, the early warning threshold calculated according to the rules is ±2.5 days, i.e., triggering an early warning when the actual completion time exceeds the range of 47.5 to 52.5 days. Warning; The quality indicator for this node is a special process adaptation pass rate of no less than 98%, and the calculated numerical difference rate warning threshold is -1%, that is, when the actual pass rate is lower than 97% (numerical difference rate is less than -1%), a warning is triggered; For the node of completing the interface connection with supplier A, the planned completion time is the 15th day of the interface connection phase, with a total planned time of 50 days, the time warning threshold is ±2.5 days, the quality indicator is an interface connectivity rate of no less than 99.5%, and the numerical difference rate warning threshold is -0.5%; Following this logic, the corresponding warning thresholds are calculated and generated for each customized progress control node, and the warning thresholds are associated with and stored with the node information to form a complete progress control and warning configuration system for this new project.

[0075] like Figure 2 As shown, embodiments of the present invention also provide an intelligent progress control system for information engineering supervision, comprising: The data acquisition module is used to collect raw progress data from the testing systems of the client, vendor, and third parties, perform data cleaning and format standardization to obtain regularized data, and store it in the central database. The dynamic optimization module is used to set multiple quality control nodes in the data processing flow based on the data stored in the central database, and to construct a multi-dimensional feature space based on the data processing quality indicators of the quality control nodes; to divide the multi-dimensional feature space into regions, identify the data feature distribution patterns of each region, and obtain data processing optimization parameters; and to dynamically optimize the data processing process using the optimization parameters to obtain optimized real-time data. The comparison module is used to automatically compare the optimized real-time data with preset key nodes and thresholds of the project plan to obtain the comparison results; The early warning module is used to make real-time early warning judgments based on comparison results. When a deviation from the schedule or failure to meet key indicators is identified, an early warning event is automatically generated, and its transmission impact on the critical path of the project is analyzed through a pre-configured rule engine to obtain early warning information. The processing module is used to push the early warning information to the systems of Party A and Party B through the message middleware and collect feedback status to complete the closed-loop collaborative processing, while updating the processing results to the central database. The optimization strategy module is used to perform pattern recognition analysis based on the central database and historical project data to extract the pattern of progress deviation, and dynamically optimize the planned node thresholds and buffer times of subsequent stages to obtain the optimization strategy. The adaptation module is used to build a configurable progress management template library based on optimization strategies, so as to quickly adapt to new types of projects and generate corresponding progress management nodes and early warning thresholds.

[0076] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0077] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An intelligent progress control method for information engineering supervision, characterized in that, The method comprises: Collecting original progress data in the first party, the second party and the third party test system, carrying out data cleaning and format standardization processing, obtaining regular data, and storing into the central database; Based on the regular data, a plurality of quality control nodes are set in the data processing process, and a multi-dimensional feature space is constructed based on the data processing quality index of the quality control nodes; the multi-dimensional feature space is regionally divided, the data feature distribution law of each region is identified, to obtain data processing optimization parameters; the data processing process is dynamically optimized by using the optimization parameters, and optimized real-time data is obtained; The optimized real-time data is automatically compared with the preset project plan key node and threshold value, and the comparison result is obtained; Based on the comparison result, real-time early warning judgment is carried out, when the progress deviation or key indicator is identified, the early warning event is automatically generated, and the conduction influence on the project critical path is analyzed through the preconfigured rule engine, and the early warning information is obtained; The early warning information is pushed to the first party and the second party system through the message middleware and the feedback state is collected to complete the closed-loop collaborative processing, and the processing result is updated to the central database; Based on the central database, combined with historical project data, mode recognition analysis is carried out to extract progress deviation law, and the plan node threshold and buffer time of subsequent stage are dynamically optimized according to the progress deviation law, and the optimization strategy is obtained; Based on the optimization strategy, a configurable progress control template library is constructed, which is used for rapid adaptation and generation of corresponding progress control node and early warning threshold in new type project. 2.The intelligent progress control method of information engineering supervision according to claim 1, characterized in that, Collecting original progress data in the first party, the second party and the third party test system, carrying out data cleaning and format standardization processing, obtaining regular data, and storing into the central database, comprising: Through a plurality of preset data interfaces, the first party management system, the second party development system and the third party test system are connected respectively, and the original progress data in the first party management system, the second party development system and the third party test system is collected in real time; The original progress data is cleaned by filling in the missing data points and correcting the abnormal data values to obtain the cleaned data; Based on the cleaned data, the heterogeneous data of different sources is converted according to the unified structured format to obtain the regular data, and the regular data is stored into the central database. 3.The intelligent progress control method of information engineering supervision according to claim 2, characterized in that, Based on the regular data, a plurality of quality control nodes are set in the data processing process, and a multi-dimensional feature space is constructed based on the data processing quality index of the quality control nodes, comprising: Based on the regular data, input verification node, conversion logic node and output calibration node quality control nodes are set in the data processing process respectively; Based on the quality control nodes, the number of valid records is counted in the input verification node, and the proportion of the number of valid records in the total number of records is calculated to obtain the data integrity ratio, the number of data entries meeting the predetermined format specification is counted in the conversion logic node, and the specification compliance ratio is calculated to obtain the format specification degree, and the matching accuracy of the output value and the standard reference value is calculated in the output calibration node to obtain the value accuracy; Based on the data integrity ratio, the format specification degree and the value accuracy, a multi-dimensional feature space is constructed.

4. The intelligent progress control method of information engineering supervision according to claim 3, characterized in that, The multi-dimensional feature space is regionally divided, and data feature distribution rules of each region are identified to obtain data processing optimization parameters; The data processing process is dynamically optimized using the optimization parameters to obtain optimized real-time data, including: The distribution density of each data point in the multi-dimensional feature space is calculated, and a high-density core region and a low-density edge region are identified according to a preset density threshold; Based on the high-density core region and the low-density edge region, the density gradient change between adjacent regions is quantitatively analyzed, and a boundary point set is formed by selecting points with peak gradient change rates; Based on the boundary point set, topological connection processing is performed on each boundary point to form a boundary range of each region; Based on the boundary range of each region, statistical analysis is performed on the distribution of data points in each boundary range to obtain the data feature distribution rules of each region; According to the data feature distribution rules, the data aggregation center position of the high-density core region is determined, the coordinate difference between the aggregation center position and a preset ideal centroid position is calculated, and an offset vector is obtained; Based on the offset vector, its spatial distance module length and angle direction relative to the coordinate axis are determined, and the module length and direction are used as optimization parameters of the data processing process; Using the optimization parameters, adjustment weights of data cleaning rules and fault tolerance thresholds of format conversion are obtained; Based on the adjustment weights and fault tolerance thresholds, parameter configuration is performed on the subsequent data processing process to form an optimized data processing flow; by executing the optimized data processing flow, optimized real-time data is obtained. 5.The intelligent progress control method of information engineering supervision according to claim 4, characterized in that, The optimized real-time data is automatically compared with preset project plan key nodes and thresholds to obtain a comparison result, including: The preset project plan is parsed to extract key node definitions and their corresponding time thresholds and quality thresholds, and a project plan structure model is established; Based on the project plan structure model, the optimized real-time data is matched and mapped according to task identification and key nodes to establish a corresponding relationship between data and plan nodes; Based on the corresponding relationship, time dimension analysis is performed on the progress data of each key node, and the difference between the actual completion time and the planned time is calculated to obtain time deviation data; Based on the time deviation data, numerical dimension analysis is performed on the quality indicators of each key node, and the difference percentage between the actual numerical indicators and the planned numerical indicators is calculated to obtain numerical difference rate data; Based on the time deviation data and the numerical difference rate data, all deviation calculation results are integrated to form the comparison result. 6.The intelligent progress control method of information engineering supervision according to claim 5, wherein, Real-time warning judgment is performed based on the comparison result, and a warning event is automatically generated when progress deviation or substandard key indicators are identified, and a pre-configured rule engine is used to analyze the conduction impact on the project critical path to obtain warning information, including: Based on the comparison result, the time deviation data is matched and mapped with the preset time tolerance threshold to obtain a deviation state identifier set, and the numerical difference rate data is compared with the corresponding quality qualification standard to obtain a comparison analysis result; Based on the deviation state identification set and the comparative analysis result, a pre-warning condition judgment is performed, and a pre-warning event is automatically generated when the time deviation exceeds a time tolerance threshold or the numerical difference rate is lower than a quality qualification standard; Based on the pre-warning event, a key path influence analysis is performed through a pre-configured rule engine to obtain a key path influence analysis result; Based on the key path influence analysis result, an influence degree of the pre-warning event on the overall progress of the project is calculated to obtain pre-warning information. 7.The intelligent progress control method of information engineering supervision according to claim 6, characterized in that, The pre-warning information is pushed to the A-party and B-party systems through a message middleware and feedback states are collected to complete closed-loop collaborative processing, and the processing result is updated to the central database, including: Based on the pre-warning information, a standardized pre-warning message is obtained through format processing by the message middleware; Based on the standardized pre-warning message, a push transmission result is obtained by pushing and transmitting the message to the specified interface of the A-party management system and the B-party development system through the message middleware; Based on the push transmission result, feedback state information returned by the A-party and B-party systems is collected, and the feedback state information is analyzed and processed to obtain a state response report; Based on the state response report, associated analysis is performed in combination with the pre-warning information to complete closed-loop collaborative processing of the pre-warning event, obtain a processing result, and update the processing result to the central database. 8.The intelligent progress control method of information engineering supervision according to claim 7, wherein, Based on the central database, historical project data is combined to perform pattern recognition analysis to extract progress deviation rules, and the plan node threshold and buffer time of the subsequent stage are dynamically optimized according to the rules to obtain an optimization strategy, including: Based on the central database, the current project progress data and historical project progress data are extracted in combination with the historical project database, and the data is processed by feature engineering to obtain a standardized feature data set; Based on the standardized feature data set, a clustering analysis method is used for pattern recognition to identify typical patterns of progress deviation to obtain a progress deviation pattern set; Based on the progress deviation pattern set, the deviation characteristics under each pattern are statistically analyzed to extract the distribution rules and influencing factors of the progress deviation to obtain a progress deviation rule analysis report; Based on the progress deviation rule analysis report, a prediction analysis method is used to predict the progress trend of the subsequent stage to obtain a prediction result; the plan node threshold and buffer time setting are dynamically adjusted according to the prediction result to obtain an optimization strategy. 9.The intelligent progress control method of information engineering supervision according to claim 8, characterized in that, Based on the optimization strategy, a configurable progress control template library is constructed to be used for rapid adaptation and generation of corresponding progress control nodes and pre-warning thresholds in new types of projects, including: Based on the optimization strategy, the plan node threshold, buffer time parameter, and pre-warning condition configuration contained therein are analyzed to obtain a standardized configuration parameter set; Based on the standardized configuration parameter set, a configurable progress control template library is constructed according to the project type classification; the progress control template library includes node definition, threshold setting, and pre-warning rule; Based on the progress control template library, when applied to a new type of project, the characteristic parameters of the new project are extracted; based on the characteristic parameters of the new project, the baseline template in the progress control template library is matched and mapped to determine the finally matched baseline template; Based on the final matched benchmark template, a customized progress control node is generated through parameter adaptation adjustment; Based on the customized progress control node, a corresponding early warning threshold is generated in combination with an early warning rule engine.

10. An intelligent progress control system for information engineering supervision, which realizes the method according to any one of claims 1 to 9, characterized in that, Comprise: The acquisition module is used for collecting original progress data in the first party, the second party and the third party test system, performing data cleaning and format standardization processing, obtaining regular data, and storing into the central database; The dynamic optimization module is used for setting multiple quality control nodes in the data processing process based on the data stored in the central database, and constructing a multi-dimensional feature space based on the data processing quality index of the quality control nodes; The multi-dimensional feature space is divided into regions, and the data feature distribution law of each region is identified to obtain data processing optimization parameters; the data processing process is dynamically optimized using the optimization parameters to obtain optimized real-time data; The comparison module is used for automatically comparing the optimized real-time data with the preset project plan key node and threshold to obtain a comparison result; The early warning module is used for real-time early warning judgment based on the comparison result, and when the progress deviation or key indicators are not up to standard are identified, an early warning event is automatically generated, and the influence on the project critical path is analyzed through the pre-configured rule engine to obtain early warning information; The processing module is used for pushing the early warning information to the first party and the second party system through the message middleware and collecting the feedback state to complete the closed-loop collaborative processing, and updating the processing result to the central database; The optimization strategy module is used for mode recognition analysis based on the central database in combination with historical project data to extract progress deviation rules, and dynamically optimizing the plan node threshold and buffer time of the subsequent stage according to the rules to obtain an optimization strategy; The adaptation module is used for constructing a configurable progress control template library based on the optimization strategy, so as to be used for quickly adapting and generating corresponding progress control nodes and early warning thresholds in new types of projects.

Images