Full-cycle engineering project management system and method

Abstract

The invention relates to the technical field of artificial intelligence and engineering project management, provides a full-cycle engineering project management system and method, and aims to solve the problems of data fragmentation, decision-making dependence on manpower and low execution collaboration efficiency in traditional engineering project management. The system comprises a data layer, a processing layer and an application layer. The data layer integrates a historical change data warehouse, dynamic market data, a supply chain collaborative database and BIM model full life cycle management, and multi-source heterogeneous data association storage is achieved. The processing layer extracts changed text key information through an NLP semantic analysis module, judges the priority, outputs a quantitative decision basis in combination with an intelligent decision support module (risk prediction, multi-source data fusion and BIM simulation), and realizes task optimal allocation and progress monitoring through an automatic execution module; and the application layer provides an interactive interface to support full-process visual collaboration. The method comprises the four stages of change request submission, NLP semantic analysis, approval decision making and automatic execution, and closed-loop management is formed.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence and engineering project management, and specifically to a full-cycle engineering project management system and method. Background Art

[0002] In the full-cycle management of engineering projects, the traditional model faces three major technical challenges that need to be solved urgently: First, data fragmentation is severe, creating information silos. Historical change records, real-time market prices, supply chain data, and BIM models are scattered across independent systems, lacking a unified linkage mechanism. This necessitates manual cross-platform data retrieval for change analysis, which is not only inefficient but also leads to incomplete decision-making due to data asynchrony. For example, estimating change costs requires manual querying of material prices and historical cases. BIM models cannot be automatically linked to construction change records, making it difficult to quickly assess the impact of changes on the structure.

[0003] Second, the decision-making process relies on manual labor and lacks intelligence. Change requests are often unstructured text, requiring manual extraction of key information, which can easily lead to omissions or misjudgments. Risk assessment relies on personal experience and lacks quantitative analysis tools, making it difficult to provide timely warnings for risks such as project delays and cost overruns. Change simulations in BIM models are isolated from other data, failing to provide multi-dimensional verification for decision-making.

[0004] Third, execution coordination is inefficient and closed-loop management is lacking. Task allocation relies on subjective assignments, making it difficult to achieve optimal matching of people with positions. Progress monitoring relies on manual reporting, which lags behind actual progress and makes it difficult to detect anomalies in a timely manner. The lack of a real-time interactive platform for cross-departmental collaboration leads to a disconnect between change execution and decision-making, creating management gaps.

[0005] Therefore, a full-cycle engineering project management system and method is proposed. Through four stages: change request submission, NLP semantic analysis, approval decision-making and automated execution, a closed-loop management is formed to solve the above problems. Summary of the Invention

[0006] Technical problems solved Traditional engineering project management has problems such as data fragmentation, manual decision-making, and inefficient execution and collaboration.

[0007] Technical Solution To achieve the above-mentioned solution, the present invention provides the following technical solutions: a full-cycle engineering project management system, applied to the field of artificial intelligence and engineering project management technology, the system includes a data layer, a processing layer and an application layer.

[0008] The data layer is used to integrate multi-source heterogeneous data and provide input data for upper-level modules, including a historical change data warehouse, a dynamic market data real-time access module, a supply chain collaborative database, and a BIM model full life cycle management module.

[0009] The processing layer is used to realize data processing and intelligent decision-making, including an NLP semantic analysis module, an intelligent decision support module and an automated execution module. The NLP semantic analysis module is used to convert unstructured change text into structured data and determine priority, including a key information extraction engine and an intelligent priority determination system. The intelligent decision support module is used for risk assessment, multi-source data fusion and BIM change simulation, including a risk prediction engine, a multi-source data fusion platform and a BIM change simulation subsystem. The automated execution module is used for task generation, allocation and progress monitoring, including a task allocation and execution system and a progress monitoring and early warning system.

[0010] The application layer is used to provide a user interaction interface, including a change submission interface, an approval decision interface, and an execution monitoring interface, supporting change request submission, approval process visualization, and real-time tracking of task progress.

[0011] Preferably, the historical change data warehouse stores complete information on change requests for past projects, including the change initiation time, applicant, change type, approval record, cost comparison, schedule impact, and risk events. It uses a relational database to store structured data, and a file system and object storage to store unstructured data, and associates structured data with unstructured data through a unique identifier.

[0012] Preferably, the dynamic market data real-time access module obtains construction material prices, equipment rental fees, and labor cost index through API interfaces and crawler technology, and stores them in a time series database after cleaning, supports querying price trends by time dimension, and provides real-time basis for change cost estimation.

[0013] Preferably, the key information extraction engine uses a Transformer-based sequence labeling model to parse the change text and extract key information such as the cause of the change, the content of the change, and the scope of impact. The model captures the dependencies between words through a multi-head attention mechanism and outputs a label sequence through a CRF layer. The priority intelligent determination system performs sentiment scoring on key information text based on an LSTM model, generates priority labels based on preset thresholds, and high-priority changes automatically trigger an expedited approval process.

[0014] Preferably, the risk prediction engine uses a random forest algorithm to build a risk assessment model, inputs change text features, project background features and historical similar change features, and outputs the risk probability distribution of construction delays, cost overruns and safety accidents; the multi-source data fusion platform establishes a weighted fusion model to process four types of data sources: market, supply chain, history and BIM, and calculates a comprehensive score to support decision-making; the BIM change simulation subsystem parses change parameters based on the IFC standard, loads the model through the BIMAPI and uses the AABB hierarchical bounding box algorithm for collision detection, and outputs three-dimensional visual simulation results.

[0015] Preferably, the task allocation and execution system generates a task list based on a preset change template, adopts the Hungarian algorithm to solve the bipartite graph matching problem to achieve optimal task allocation, and the objective function is the comprehensive cost of task allocation, that is, the lowest score of time cost and ability matching; the progress monitoring and early warning system adopts the dynamic time warping DTW algorithm to match the actual progress with the planned progress, triggers a graded early warning when the cumulative difference exceeds a threshold, and automatically pushes exception handling suggestions.

[0016] A full-cycle engineering project management method, which is applied to the above-mentioned full-cycle engineering project management system, comprises the following steps: S1. Change request submission phase: Change requests entered by users are received in voice, text, and image formats. Preliminary verification is performed to ensure data integrity. The text is pre-processed to include word segmentation, stop word removal, and terminology standardization.

[0017] S2.NLP semantic analysis phase: The Transformer model is used to extract key information from the change text, including the cause of the change, content, and scope of impact. The LSTM model is used to perform sentiment scoring and determine priority, and the extracted results are encapsulated as structured change objects.

[0018] S3. Approval and decision-making stage: Start the risk prediction model to output a three-dimensional risk probability matrix, calculate the comprehensive score through the multi-source data fusion platform, combine the BIM change simulation results to generate decision recommendations, and support manual approval and correction.

[0019] S4. Automated execution phase: Generate a task list based on the change type and matching the preset template. Use the Hungarian algorithm to assign tasks. Use RPA tools to automatically execute repetitive operations. Monitor task progress in real time and trigger exception warnings to form a closed-loop management system.

[0020] Preferably, the BIM change simulation in step 3 includes: parsing the geometric parameters and attribute parameters in the change content, loading the corresponding version model through BIMAPI and locating the affected components, using the AABB hierarchical bounding box algorithm to perform collision detection, and displaying the detection results in a three-dimensional visual form.

[0021] Preferably, the task list in step 4 includes task nodes, responsible positions and duration requirements, and the dependency relationship between task nodes is visualized. When the task progress delay exceeds a threshold, the system automatically adjusts the subsequent task time nodes and pushes resource allocation suggestions.

[0022] Preferably, the comprehensive score calculation method of the multi-source data fusion platform in step 3 is:

[0023] in, are the weights of market data, supply chain data, historical data, and BIM data respectively. The standardized score range of 0-100 is for each data source, which is calculated based on price fluctuations, supply capacity, impact of historical changes and complexity of model modification.

[0024] Beneficial effects Compared with the prior art, the present invention provides a full-cycle engineering project management system and method, which has the following beneficial effects: 1. In this solution, the full-cycle engineering project management system breaks down information silos through the deep integration of multi-source heterogeneous data (historical change data, market data, supply chain data and BIM models), achieving the association and reuse of full-cycle data; with the help of NLP semantic analysis, risk prediction engine and BIM simulation intelligent engine, it transforms unstructured information into quantitative decision-making basis, replacing traditional manual analysis and experience judgment, significantly improving the scientific nature and efficiency of decision-making; through the automated execution module and cross-departmental collaborative interface, it realizes closed-loop management of task allocation, progress monitoring and exception handling, greatly reducing manual coordination costs, and overall improving the accuracy, coordination and intelligence level of full-cycle management of engineering projects.

[0025] 2. In this solution, the full-cycle engineering project management method uses NLP semantic analysis to achieve structured extraction and intelligent priority determination of change texts, replacing manual organization and subjective grading to ensure the consistency and timeliness of information processing; with the help of multi-source data fusion evaluation, risk quantification analysis and BIM simulation verification, it provides a multi-dimensional scientific basis for approval decisions, avoiding the limitations of traditional experience-based decision-making; through automated task allocation, progress monitoring and abnormal adjustment mechanisms, it achieves precise control and dynamic optimization of the entire change execution process, forming a complete closed loop of "request-analysis-decision-execution", which significantly improves the process standardization, decision reliability and execution efficiency of engineering project management. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Schematic diagram of the overall system architecture of the present invention; Figure 2This is a schematic diagram of the data layer architecture of the present invention; Figure 3 This is a schematic diagram of the processing layer architecture of the present invention; Figure 4 Schematic diagram of the application layer architecture of the present invention. DETAILED DESCRIPTION

[0027] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0028] See also Figures 1 to 4 The present invention proposes a full-cycle engineering project management system and method, including the following contents: 1. Full-cycle engineering project management system (1) Data layer: the integration hub for multi-source heterogeneous data As the foundation of system operation, the data layer undertakes the core tasks of data collection, cleaning, storage and management. Its core role is to provide high-quality input data for upper-layer modules. This layer integrates four key data categories: 1. Historical Change Data Warehouse This database stores complete information on all change requests from past projects, including the change initiation time, applicant, change type (design / construction / procurement, etc.), change description, approval process records, implementation process data, pre- and post-change cost comparisons, schedule impacts, and risk event records. For example, the historical data for a bridge project includes over 300 pile foundation design change records, each of which is associated with attachments such as geological survey reports, BIM model versions, and construction logs, resulting in a mixed storage of structured and unstructured data. The data format uses a relational database (such as PostgreSQL) to store structured data, and a file system or object storage (such as MinIO) to store unstructured data such as drawings and documents, with associations achieved through unique identifiers.

[0029] 2. Real-time access to dynamic market data APIs are used to obtain real-time market data on construction material prices (such as steel, concrete, and pipes), equipment rental costs, and labor cost indices. For example, by connecting to commodity trading platforms (such as Shanghai Steel Union), steel price fluctuation data is updated every 15 minutes. Quarterly wage guidelines are captured using crawler technology. After data cleansing, it is stored in a time-series database (such as InfluxDB), supporting time-based price trend queries and providing real-time data for change cost estimation.

[0030] 3. Supply Chain Collaboration Database Integrate supplier information (qualifications, production capacity, historical cooperation evaluations), inventory data (raw material inventory, turnover cycle), logistics track (transportation route, timeliness, abnormal events), etc. By connecting with the ERP system, real-time information such as material purchase order status and supplier delivery date can be obtained. For example, the supply chain database of a subway project records the concrete supply capacity of more than 50 suppliers. When changes lead to an increase in concrete demand, the system automatically evaluates the remaining capacity of each supplier (for example, supplier A has a daily output of 300m3). 3 , the current order occupies 200m 3 ) and the shortest delivery time (e.g. Supplier B can deliver within 48 hours), providing a quantitative basis for resource allocation.

[0031] 4. BIM model full life cycle management The 3D BIM model of a project is stored, including specialized models for architecture, structure, electromechanical engineering, and other disciplines, as well as model version information (e.g., different versions for the design phase, construction phase, and completion phase). Each model component is associated with geometric parameters (dimensions, coordinates), attribute parameters (material type, strength grade), and associated documents (design drawings, calculations). For example, the BIM model of a commercial complex contains over 100,000 components, each with a unique UUID identifier. Detailed component information (e.g., 800mm air duct diameter, galvanized steel material) can be retrieved through an API, providing an accurate 3D data foundation for change simulation.

[0032] (2) Processing layer: the core engine of intelligent management The processing layer is the "brain" of the system and consists of three core modules. The modules work together through data interfaces to complete the entire intelligent operation process from data processing to decision output.

[0033] 1.NLP semantic analysis module: intelligent parser for changing text The core function of this module is to convert unstructured change request text into structured data, extract key information and evaluate priorities. It specifically includes two sub-functions: (1) Key information extraction engine Adopting Transformer-based sequence tagging models (such as BERT-CRF) to achieve fine-grained analysis of the changed text. The input layer converts the text into a sequence of word vectors after word segmentation. ,in It is The word embedding (WordPiece) of each word is represented by a dimension of 768 (consistent with the BERT-base model). The model captures the dependencies between words through a multi-head attention mechanism (12 attention heads). For example, in the sentence "foundation settlement leads to insufficient pile bearing capacity, and the pile length needs to be increased by 2 meters", the causal relationship between "foundation settlement" and "pile bearing capacity" is reflected through the attention weight. The output layer uses the CRF layer to predict the label sequence, and the label set is defined as ,Each word corresponds to a label, such as “foundation settlement” is labeled as “change reason”, and “increase pile length by 2 meters” is labeled as “change content”.

[0034] During model training, a large-scale annotated dataset (containing more than 200,000 historical change texts) was used for fine-tuning, and the loss function was cross-entropy loss:

[0035] in, is the text length (number of words), is the number of label categories (4 categories), is the true label (0 or 1, 1 means the i-th word belongs to the j-th label), The model predicts the probability that the i-th word belongs to the j-th category label. According to actual tests, the F1 value of key information extraction can reach over 92%.

[0036] (2) Priority intelligent determination system The LSTM-based sentiment analysis model assesses the urgency of the change. The input is the text vector concatenated from the word vectors of key information. The model captures long-range semantic dependencies through a three-layer LSTM network (256 hidden units per layer). Determine how much historical information to retain, input gate Control the new information of the current input, the output gate Regulating cell state Generate hidden state , the specific formula is as follows:

[0037] in, represents the time step, The word vector input for the current time step; 、 、 、 are the weight matrices of the forget gate, input gate, cell state update, and output gate respectively; 、 、 、 is the corresponding bias vector; is the sigmoid function (output range 0-1), is the hyperbolic tangent function (output range -1-1); Represents element-wise multiplication. The LSTM model outputs sentiment scores , the higher the score, the more urgent the change. Preset threshold (high priority), (Medium priority), for example, the sentiment score of "The risk of tunnel mud burst requires immediate adjustment of the support plan" is 0.8, which is automatically marked as high priority, triggering an expedited approval process.

[0038] 2. Intelligent Decision Support Module: Decision-making Hub for Multi-Dimensional Analysis This module integrates three major functions: risk assessment, multi-source data fusion, and BIM simulation, providing all-round support for approval: (1) Risk prediction engine The random forest algorithm is used to build a risk assessment model. The input features include: Change text features: keyword weights extracted through TF-IDF, sentiment scores of key information Project background characteristics: project stage (design / construction / operation and maintenance), contract type (EPC / DBB), investment scale Characteristics of historical similar changes: frequency of risk occurrence and average impact duration of similar changes A single decision tree selects the split attribute through Gini impurity, and finally integrates the prediction results of T=100 decision trees through voting method to output the risk probability distribution , corresponding to the probabilities of construction delays, cost overruns, and safety accidents, respectively. For example, the design change prediction results for a building construction project are: a 35% probability of construction delay (threshold 20%) and a 28% probability of cost overrun (threshold 25%), triggering a risk warning.

[0039] (2) Multi-source data fusion platform A weighted fusion model was established to process n=4 data sources (market, supply chain, history, BIM). The formula is:

[0040] Among them, the weight Determined by data importance analysis (such as market price , supply chain , historical data , BIM simulation ); Score of each data source Normalized to the interval [0,100]: Market Score : Based on the product of price fluctuation and the amount of change (e.g., if the steel price increases by 5% and the amount of change is 10%, the score is 80) Supply Chain Score : Based on supply capacity (remaining capacity ≥ demand = 90) and delivery time (≤ 24 hours = 90) Historical scores : Refer to the cost overrun rate (≤10% = 85) and construction delay rate (≤5 days = 85) of similar changes BIM score : Based on the collision detection results (no collision, 80) and the model modification complexity (component adjustment amount ≤ 5%, 75) For example, the fusion score for a drainage pump change is: .

[0041] (3) BIM change simulation subsystem Implement the interaction between change information and BIM models based on the IFC standard. Specific steps: Analyze the geometric parameters (such as component size and position adjustment) and property parameters (material replacement) in the changed content Load the corresponding version of the model through BIMAPI (such as AutodeskForge) and locate the affected component set The AABB hierarchical bounding box algorithm is used for collision detection. The formula is:

[0042] in, , is the bounding box of the changed component, is the bounding box of the adjacent components, Indicates the maximum coordinate of the bounding box A on the d axis (x / y / z), The minimum coordinate of bounding box B on the d-axis is shown. The detection results are displayed in 3D visualization, with collision areas highlighted in red. For example, the drainage pump (500mm diameter) does not collide with the existing pipeline (1500mm spacing), and the simulation report indicates "sufficient installation space."

[0043] 3.Automated execution module This module automates the entire process from task generation to progress monitoring and includes two core functions: (1) Task allocation and execution system Task list generation is based on a template engine, with over 10 pre-set change templates (such as design change templates and construction plan change templates). Each template defines task nodes (such as drawing modification, budget adjustment, and on-site briefing), responsible positions (designer, cost estimator, construction manager), and deadline requirements (complete drawing approval within 3 days). Task allocation uses the Hungarian algorithm to solve the bipartite graph matching problem, with the objective function being:

[0044] in, is the number of tasks, is the number of responsible persons, The comprehensive cost (time cost + ability matching score) of assigning task i to person j is is a decision variable (1 indicates allocation, 0 indicates no allocation). The optimal allocation solution is solved using a matrix inversion algorithm. For example, assigning the "drawing modification" task to Engineer Zhang, whose design ability is 90 points, and assigning the "budget adjustment" task to Engineer Li, whose cost estimation ability is 95 points.

[0045] (2) Progress monitoring and early warning system Use the Dynamic Time Warping (DTW) algorithm to match the actual progress with the planned progress, with the input being a discrete time series and ,in is the actual completion percentage, is the planned completion percentage. Distance matrix The recursive calculation is as follows:

[0046] in, The difference between actual and planned progress (e.g. ), is the cumulative distance. When the threshold is set to 0.3, three levels of warning are triggered: yellow warning (delay of 1-3 days), orange warning (delay of 3-7 days), and red warning (delay of more than 7 days). For example, if Task 3 (procurement of drainage pumps) is reported as "supplier out of stock" on the second day, the system detects a 50% delay in progress, triggering a yellow warning and automatically switching to an off-site warehouse for stocking.

[0047] (3) Application layer: unified entry point for user interaction The application layer uses a microservices architecture to develop web and mobile apps, providing three core functional interfaces: 1. Change submission interface The system supports rich text input, attachment upload (drawings, inspection reports), and BIM model preview, automatically linking historical similar change cases to help users refine change descriptions. For example, when entering "pile foundation," a list of similar changes pops up, displaying the treatment plans and impact data of past pile foundation changes (e.g., "In 2023, the pile foundation of a certain project was lengthened by 2 meters, increasing costs by 120,000 yuan").

[0048] 2. Approval decision interface Visualize the NLP analysis results (key information highlighted), risk assessment reports (probability distribution histograms), and BIM clash detection results (3D model comparison views). The approver can adjust the weights of each factor through a slider and view the changes in decision-making suggestions in real time, supporting annotations and electronic signatures.

[0049] 3. Execution monitoring interface Display the task progress in the form of a Gantt chart, update the completion status of the responsible person in real time (such as "Engineer Zhang - Drawing modification: 80% completed"), integrate IoT device data (such as the GPS location of construction machinery and sensor data), automatically highlight abnormal status in red and display the detailed reasons (such as "Steel not delivered on time - Supplier transportation delay"). Visualize task dependencies (such as Task 4 starts after Task 1 is completed), intuitively showing the process connection.

[0050] II. Full-cycle engineering project management method (I) Change request submission stage: Transformation from text to structured data 1. User input The construction worker / designer submits a change request through the mobile APP or the Web端, supporting voice input (automatically converted to text) and photo recognition (extracting drawing change marks). For example, the construction worker discovers geological inconsistencies on site, takes a photo of the borehole log and describes it verbally: "The thickness of the silty clay at K1+200 is 2 meters less than the exploration report, and the pile length needs to be increased."

[0051] 2. Preliminary verification The system automatically checks whether the required fields (such as the change location and impact scope) are complete and whether the attachment format is compliant (only supports PDF, DWG, IFC). If missing, it prompts the user to supplement (such as "Please fill in the expected impact on the construction period of the change"), ensuring the integrity of the submitted data.

[0052] 3. Text preprocessing Perform word segmentation, stop word removal, simplified and traditional Chinese conversion, etc. on the input text to generate a pure word sequence, providing a basis for subsequent NLP analysis. For example, convert "砼" to "concrete", remove meaningless words such as "的" and "了", and unify the term expressions.

[0053] (II) NLP semantic analysis stage: Intelligent extraction and priority determination of key information 1. Key information extraction The Transformer model analyzes the text word by word, outputs the label probability of each word, and decodes the optimal label sequence through the Viterbi algorithm. For example, for the input text "Due to heavy rain, the foundation pit is flooded, and 2 drainage pumps need to be added, with the expected construction period extended by 3 days", the extraction result is: Reason for change: Heavy rain caused the foundation pit to be flooded Change content: Add 2 drainage pumps Impact: Construction period extended by 3 days 2. Sentiment Analysis and Classification The LSTM model assigns sentiment scores to key information and generates priority tags based on pre-set rules. If the sentiment score S = 0.7 (> α = 0.6), the message is marked as high priority, the system automatically adds an "urgent" flag, and notifies the lead engineer via text message.

[0054] 3. Data Encapsulation The extracted key information, priority, original text, attachments, etc. are encapsulated into a change object in JSON format, transmitted to the intelligent decision support module through the API interface, and stored in the historical change database to generate a unique change ID (such as BC-20231015-001).

[0055] (III) Approval and Decision-Making Stage: Multi-dimensional Analysis Supports Scientific Decision-Making 1. Risk assessment initiated The risk prediction model loads the text features of the change object, basic project data (such as a contract duration of 180 days and a budget of 50 million yuan), and historical similar change data (the average duration of 30 drainage system changes in the past year affected the project by 2 days), and outputs a three-dimensional risk probability matrix. (Construction period / cost / safety risk).

[0056] 2. Multi-source data fusion calculation The market data API was simultaneously called to obtain the current steel price fluctuation (+5%), and the supply chain system feedback that the drainage pump inventory was sufficient (delivery within 24 hours). Combined with the cost impact of similar changes in historical change data (average overrun of 15%) and the model modification complexity of BIM simulation (component adjustment amount ≤ 5%), a comprehensive impact score was calculated:

[0057] 3. BIM change simulation Analyze the equipment parameters in the change (drainage pump dimensions: 500mm diameter, 1200mm height), locate the installation position in the foundation pit model, and check for collisions with existing pipelines (300mm diameter, 1500mm spacing): Horizontal spacing: 1500mm ≥ 500mm + 300mm (safety distance), no collision Longitudinal height: pit depth 5m ≥ 1200mm (equipment height), sufficient installation space Output simulation report: no collision risk, installation space meets requirements.

[0058] 4. Decision suggestion generation Based on the risk assessment (construction period risk exceeds the threshold), comprehensive score (M=81.25>70), and BIM simulation (no collision), the system generates a suggestion: "Agree to the change, pay attention to the trial run inspection of the drainage pump after installation", and attaches detailed basis (risk report, BIM screenshots, market quotation).

[0059] 5. Manual approval and revision Approving personnel browse the system suggestions. If they have any questions, they can manually trigger a request for additional information (such as requesting geological radar scanning data) or directly modify the decision (such as rejecting the change due to funding budget constraints). The approval results are fed back to the system in real time.

[0060] (IV) Automated Execution Phase: From Task Allocation to Closed-Loop Management 1. Task list generation According to the change type (construction plan change), match the preset template and generate a checklist containing 5 task nodes: Task 1: Design Department modifies drainage system drawings (Responsible Person: Engineer Zhang, Duration: 2 days) Task 2: Contract Department Budget Adjustment (Responsible Person: Engineer Li, Duration: 1 day) Task 3: Materials Department purchases drainage pumps (Responsible Person: Engineer Wang, Duration: 3 days) Task 4: On-site briefing by the Engineering Department (Responsible Person: Engineer Zhao, Duration: 0.5 days, dependent on completion of Task 1) Task 5: Quality Department Acceptance (Responsible Person: Zhou Gong, Duration: 0.5 days, Depends on Task 3) 2. Intelligent task allocation The Hungarian algorithm calculates the optimal allocation plan based on the responsible person's ability matrix (Engineer Zhang: design ability 90, coordination ability 80; Engineer Li: cost estimation ability 95, communication ability 75) to ensure the lowest total task cost. For example, the drawing modification task is assigned to Engineer Zhang, and the budget adjustment task is assigned to Engineer Li.

[0061] 3. RPA Automation Execution Robotic process automation tools automate repetitive operations: Extract the drawing number from the changed object, automatically update the CAD drawing version number and upload it to the document management system; Create a new purchase requisition in the ERP system based on the budget adjustment amount (+50,000 yuan); Generate change notification emails and automatically copy all relevant parties (owners, supervisors, and subcontractors).

[0062] 4. Real-time progress monitoring The system synchronizes the status of each task every hour: Task 1: At 16:00 on the first day, Mr. Zhang marks "Drawing modification completed", and the system automatically triggers the start condition of Task 4. Task 3: On the second day, Engineer Wang reported that the supplier had notified him that the drainage pumps were out of stock. The system detected a 50% delay in progress (planned completion: 66%, actual completion: 16%), triggering a yellow alert and prompting the Materials Department to initiate the supplier selection process (automatically switching to an off-site warehouse for delivery, with an estimated one-day delay in delivery). 5. Exception handling and closed loop To address the progress delay, the project manager initiated a coordination meeting on the monitoring interface and adjusted the procurement plan to "transfer goods from an off-site warehouse, with an estimated arrival time extension of one day." The system automatically updated Task 3's duration (from three days to four days) and recalculated the timelines for subsequent tasks (Task 5's start time was delayed by one day). After the change was implemented, the Quality Department uploaded the acceptance report (with a 100% pass rate), and the system automatically archived all relevant data (change requests, approval records, and execution logs), forming a complete change management closed loop.

[0063] Finally, it should be noted that the above embodiments are merely examples for the purpose of illustrating the present invention and are not intended to limit the embodiments. Those skilled in the art will readily appreciate that other variations or modifications based on the above description are possible. It is not necessary and impossible to provide an exhaustive list of all embodiments. However, obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A full-cycle engineering project management system, applied in the field of artificial intelligence and engineering project management technology, characterized by: The system includes a data layer, a processing layer and an application layer; The data layer is used to integrate multi-source heterogeneous data and provide input data for upper-layer modules, including historical change data warehouse, dynamic market data real-time access module, supply chain collaboration database and BIM model full life cycle management module; The processing layer is used to implement data processing and intelligent decision-making, including an NLP semantic analysis module, an intelligent decision support module and an automated execution module. The NLP semantic analysis module is used to convert unstructured change text into structured data and determine priority, including a key information extraction engine and an intelligent priority determination system. The intelligent decision support module is used for risk assessment, multi-source data fusion and BIM change simulation, including a risk prediction engine, a multi-source data fusion platform and a BIM change simulation subsystem. The automated execution module is used for task generation, allocation and progress monitoring, including a task allocation and execution system and a progress monitoring and early warning system. The application layer is used to provide a user interaction interface, including a change submission interface, an approval decision interface, and an execution monitoring interface, supporting change request submission, approval process visualization, and real-time tracking of task progress.

2. A full-cycle engineering project management system according to claim 1, characterized in that: The historical change data warehouse stores complete information on change requests for past projects, including the change initiation time, applicant, change type, approval record, cost comparison, schedule impact, and risk events. It uses a relational database to store structured data and a file system and object storage to store unstructured data, and associates structured data with unstructured data through a unique identifier.

3. A full-cycle engineering project management system according to claim 1, characterized in that: The dynamic market data real-time access module obtains construction material prices, equipment rental fees, and labor cost index through API interfaces and crawler technology, and stores them in a time series database after cleaning. It supports querying price trends by time dimension and provides a real-time basis for change cost estimation.

4. A full-cycle engineering project management system according to claim 1, characterized in that: The key information extraction engine uses a Transformer-based sequence labeling model to parse the change text and extract key information such as the cause of the change, the content of the change, and the scope of impact. The model captures the dependencies between words through a multi-head attention mechanism and outputs a label sequence through the CRF layer. The priority intelligent determination system uses an LSTM model to perform sentiment scoring on key information text and generates priority labels based on preset thresholds. High-priority changes automatically trigger an expedited approval process.

5. The full-cycle engineering project management system according to claim 1, characterized in that: The risk prediction engine uses a random forest algorithm to build a risk assessment model, inputs change text features, project background features and historical similar change features, and outputs the risk probability distribution of construction delays, cost overruns and safety accidents; the multi-source data fusion platform establishes a weighted fusion model to process four types of data sources: market, supply chain, history and BIM, and calculates a comprehensive score to support decision-making; the BIM change simulation subsystem parses change parameters based on the IFC standard, loads the model through the BIMAPI, and uses the AABB hierarchical bounding box algorithm for collision detection, and outputs three-dimensional visual simulation results.

6. A full-cycle engineering project management system according to claim 1, characterized in that: The task allocation and execution system generates a task list based on a preset change template and uses the Hungarian algorithm to solve the bipartite graph matching problem to achieve optimal task allocation. The objective function is the comprehensive cost of task allocation, that is, the minimum time cost and ability matching score; The progress monitoring and early warning system uses the dynamic time warping (DTW) algorithm to match the actual progress with the planned progress. When the cumulative difference exceeds a threshold, a graded early warning is triggered and abnormal handling suggestions are automatically pushed.

7. A full-cycle engineering project management method, characterized by: The method is applied to a full-cycle engineering project management system according to any one of claims 1 to 6 above, and the method comprises the following steps: S1. Change Request Submission Stage: Receive change requests entered by users in voice, text, and image formats, perform preliminary verification to ensure data integrity, and pre-process the text for word segmentation, stop word removal, and terminology standardization. S2. NLP semantic analysis: The Transformer model is used to extract key information from the change text, including the cause, content, and scope of impact. The LSTM model is used to perform sentiment scoring and determine priority, and the extracted results are encapsulated as structured change objects. S3. Approval and decision-making stage: The risk prediction model is activated to output a three-dimensional risk probability matrix. A comprehensive score is calculated using a multi-source data fusion platform. Decision recommendations are generated based on BIM change simulation results, supporting manual approval and revision. S4. Automated execution phase: Generate a task list based on the change type and matching the preset template. Use the Hungarian algorithm to assign tasks. Use RPA tools to automatically execute repetitive operations. Monitor task progress in real time and trigger exception warnings to form a closed-loop management system.

8. A full-cycle engineering project management method according to claim 7, characterized in that: The BIM change simulation in step 3 includes: parsing the geometric parameters and attribute parameters in the change content, loading the corresponding version model through BIMAPI and locating the affected components, using the AABB hierarchical bounding box algorithm to perform collision detection, and displaying the detection results in a three-dimensional visual form.

9. A full-cycle engineering project management method according to claim 7, characterized in that: The task list in step 4 includes task nodes, responsible positions and duration requirements, and the dependency relationships between task nodes are visualized. When the task progress delay exceeds the threshold, the system automatically adjusts the subsequent task time nodes and pushes resource allocation suggestions.

10. A full-cycle engineering project management method according to claim 7, characterized in that: The comprehensive score calculation method of the multi-source data fusion platform in step 3 is: in, are the weights of market data, supply chain data, historical data, and BIM data respectively. The standardized score range of 0-100 is for each data source, which is calculated based on price fluctuations, supply capacity, impact of historical changes and complexity of model modification.

Images