Engineering construction digital project management method and system

Abstract

The invention relates to the technical field of engineering progress management, in particular to an engineering construction digital project management method and system, and the method comprises the steps: database building, real-time data collection, and model building: generating a visual construction progress model; progress deviation judgment: calculating progress deviation and performing judgment; deviation analysis: calculating a resource gap of an affected process, and generating a resource allocation priority list; resource adjustment: pushing an adjustment instruction to the construction terminal according to the priority list; simulation: simulating the adjusted construction progress, and if the deviation is not eliminated, executing a redistribution step; redistribution: executing the deviation analysis step again; and optimization: optimizing subsequent project progress plan generation logic. The system comprises a database building module, a real-time data acquisition module, a model building module, a progress deviation judgment module, a deviation analysis module, a simulation module, a redistribution module and an optimization module. The method and the device have the effect of facilitating fine management of the engineering project.

Description

Technical Field

[0001] The present application relates to the technical field of engineering progress management, and in particular to a method and system for digital project management of engineering construction. Background Art

[0002] Currently, in the construction industry, project progress management is a core component for ensuring on-time delivery, controlling costs, and optimizing resource allocation. Traditional progress management relies heavily on manually compiled tools like Gantt charts and the critical path method. Data updates lag and rely on empirical judgment, making it difficult to adapt to the dynamic changes on the construction site. With the widespread adoption of technologies like BIM (Building Information Modeling), IoT (Internet of Things), and GIS (Geographic Information Systems), project management is gradually transitioning to digitalization. By integrating multi-source data, project progress visualization, dynamic resource scheduling, and risk warnings can be achieved.

[0003] Among existing technologies, Chinese patent application publication number CN117035691A discloses a digital project management method and system for engineering construction based on BIM technology. However, this method lacks full-lifecycle data integration, making it difficult to detect and correct deviations between actual construction progress and planned progress. Resource scheduling relies on manual experience, which can easily lead to idle equipment or shortages. Furthermore, the efficiency of multi-role collaboration is low, and data-driven dynamic decision-making capabilities are weak, hindering the refined management of engineering projects. Summary of the Invention

[0004] In order to facilitate the refined management of engineering projects, this application provides a digital project management method and system for engineering construction.

[0005] In the first aspect, this application provides a digital project management method for engineering construction, which adopts the following technical solutions: A digital project management method for engineering construction includes the following steps: Database construction: Automatically extract the bill of quantities based on the BIM model and associate it with the construction schedule and real-time collected construction log data to build a dynamically updated project progress database; Real-time data collection: Real-time collection of labor distribution, equipment operation status and material consumption data at the construction site, and synchronization with the progress database; Model building: Combine the BIM model with the construction plan to conduct 4D progress simulation and generate a visual construction progress model that includes the time dimension; Progress deviation judgment: Compare the deviation between the planned progress and the actual construction progress in the visual construction progress model. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Deviation analysis: Deviation analysis is performed based on schedule deviations. Based on the results of the deviation analysis, the resource gap of the affected processes is dynamically calculated through an algorithm, and the available equipment, personnel, and material inventory data are matched to generate a resource allocation priority list; Resource adjustment: push adjustment instructions to the construction terminal according to the priority list, and update the resource occupancy status to the database; Simulation: Iterate the simulation of the adjusted construction schedule to verify the effectiveness of the resource allocation plan. If the deviation is not eliminated, perform the reallocation step; Reassign: Perform deviation analysis again and follow the deviation analysis steps; Optimization: Store historical progress data, deviation reasons, and adjustment records in the database, and use machine learning models to optimize the schedule generation logic for subsequent projects.

[0006] By adopting the above technical solutions, BIM automatically extracts the bill of quantities, associates it with the construction plan and real-time log, and forms a cross-stage data pool; multi-dimensional data synchronization: IoT collects human, machine and material data in real time to eliminate information lag; 4D model mapping: BIM+time dimension realizes progress visualization deduction; threshold warning and iterative optimization: deviation triggers dynamic resource allocation, and machine learning continuously optimizes plan generation.

[0007] This system connects the design, planning, and execution data chains, resolving data gaps caused by traditional segmented management (e.g., disconnection between BIM models and on-site progress). Sensor data updates within seconds, improving response times from traditional daily / weekly reports to minute-level responses. For example, equipment failures can instantly trigger resource reallocation. A closed-loop process encompassing early warning, analysis, adjustment, simulation, and re-optimization forms a digital version of the Plan-Do-Check-Act (PDCA) process, improving efficiency compared to existing post-correction methods. Machine learning transforms historical deviations (e.g., delays caused by the rainy season) into constraints for subsequent plans, digitizing experience. Through analysis and identification of deviations, the causes are identified and resource allocation is implemented accordingly, reducing the progress of a single construction line and allowing for simultaneous construction of multiple lines to maintain completion times. Simulations are performed before resource allocation, and a resource allocation priority table is generated based on project importance to facilitate prioritized allocation. Resource scheduling is based on data analysis, minimizing equipment idleness or shortages, improving the efficiency of multi-role collaboration, and enabling data-driven dynamic decision-making, ultimately leading to refined project management.

[0008] Optionally, the deviation analysis step specifically includes: constructing a process resource demand prediction model using a random forest algorithm, inputting historical progress data, current resource gaps and environmental variables, and outputting resource gap priority scores for the affected processes; Match available resources with the gap list to generate a resource allocation plan that minimizes scheduling costs; The allocation plan is verified with the cloud-based material inventory data in real time. If the inventory is insufficient, the supplier collaboration interface is automatically triggered to generate an expedited purchase order.

[0009] By adopting the above technical solutions, random forest can handle multivariate coupling problems (such as equipment failure affecting both concrete pouring and rebar binding), accurately identify key resource gaps, and reduce prediction errors compared to linear regression models; it can transform resource gaps from "qualitative judgment" to "quantitative scoring". For example, when cranes and welders are lacking at the same time, priority is given to replenishing resources that have a greater impact on the critical path; through API docking with supplier systems, a three-level linkage of "site-inventory-supply chain" is achieved, and the purchase order generation time is compressed from hours to minutes, reducing the risk of downtime due to waiting for materials; the random forest algorithm improves the accuracy of resource demand prediction, automatic procurement reduces the risk of downtime due to waiting for materials, and the scheduling cost minimization model reduces resource waste.

[0010] Optionally, after the progress deviation determination step, the method further includes an associated node intelligent scheduling step; The steps of intelligent scheduling of associated nodes include: data extraction step, association degree calculation step, interference evaluation step and dynamic start step; Data extraction: Extract task nodes with logical dependencies from the WBS structure tree of the BIM model and construct a process association network diagram, where nodes represent construction tasks and edges represent the preceding / successive relationships between processes; Correlation calculation: Analyze the centrality of each node in the correlation network, combine the actual impact frequency in historical construction data, and calculate the global correlation score of the node; Interference assessment: Real-time monitoring of progress deviations of the currently affected process, inferring the potential interference probability of the affected process on the associated nodes through Bayesian networks, and marking the associated nodes as eligible for parallel startup if the interference probability is lower than a preset threshold. Dynamic startup: Push instructions to the construction teams corresponding to the nodes that can be started in parallel, allocate idle labor and equipment to the node in advance, lock the inventory reserves of the required materials, generate parallel construction permits and synchronize them to the supervision system.

[0011] By adopting the above technical solutions, task dependencies are constructed based on the WBS structure tree of BIM, and the potential impact probability of progress deviation on related nodes is calculated; idle resources are allocated to nodes with low interference probability; the traditional critical path method only identifies explicit logical relationships, while this solution captures implicit dependencies (such as hidden conflicts of shared tower cranes) through process network diagrams to reduce the risk of parallel construction; the Bayesian network combines historical data (such as the probability of rework caused by changes to a certain type of node) to dynamically calculate the interference probability, which reduces the error rate compared to the static rule base; through the "Parallel Construction Permit" system, idle resources in traditional serial processes are The idle tower cranes and work teams are flexibly allocated to non-critical paths, which improves equipment utilization and ensures that construction is completed in an orderly and timely manner; parallel construction opportunities are identified through the association network to shorten the critical path; the Bayesian network quantifies the probability of interference to avoid risks caused by blind parallel construction; idle resources are reallocated to improve utilization. At the same time, through the mobilization and allocation of resources, multi-node joint construction can alleviate the delays caused by single-line construction. Through the construction location, node association is established to facilitate simultaneous multi-line construction, reduce the impact of current main line construction delays on the construction period, and at the same time reduce resource waste, thereby saving costs.

[0012] Optionally, the associated node intelligent scheduling step further includes a priority sorting step: Prioritization includes: conflict detection step, multi-objective optimization step and elastic buffer step; Conflict detection: When multiple associated nodes meet the conditions for parallel startup, a resource-time conflict matrix is constructed to quantify the duration of each node's occupation of key resources and the spatial overlap area; Multi-objective optimization: Establish a mixed integer programming model with the goal of maximizing the duration reduction rate and minimizing resource conflicts. Input the conflict matrix and current resource distribution data, and output the optimal start sequence. Elastic buffering: Dynamic buffering time windows are allocated to the highest-priority nodes. If their actual progress lags behind the predicted value, resource withdrawal instructions are automatically triggered to reduce interference with the main critical path.

[0013] By adopting the above technical solution, a resource-time conflict matrix is constructed when multiple nodes can run in parallel. A mixed integer programming model is used to output the optimal startup sequence, allocating flexible buffer time windows to high-priority nodes. The multi-node resource competition problem is converted into a matrix operation. For example, the conflicting usage periods of tower cranes at nodes A and B can be quantified and ranked. Mixed integer programming balances duration reduction (e.g., starting node X early can save five days) with resource conflict costs (e.g., additional rental fees) to find the most economically optimal solution. The buffer time window dynamically expands and contracts based on actual progress (compressing the buffer period if a node completes early). Compared to a fixed buffer period, this saves redundant time to quantify resource conflicts, avoid confusion caused by multi-node running in parallel, and balance the contradiction between duration reduction and resource conflicts. The buffer time window enhances scheduling fault tolerance.

[0014] Optionally, before the dynamic startup step, a risk hedging step is also included: The risk hedging steps include: scenario simulation step, impact diffusion analysis step and hedging strategy generation step; Scenario simulation: Before starting the associated nodes, the digital twin engine simulates the entire construction process to predict possible quality defects and safety risks. Impact Diffusion Analysis: Build a cascading failure model based on complex networks to quantify the propagation path and probability of the impact of construction anomalies at associated nodes on the overall construction period; Hedging strategy generation: If the simulation results show that the risk diffusion value exceeds the tolerance, the node startup instruction will be frozen and alternative plans will be generated until the risk value drops to a safe threshold.

[0015] By adopting the above technical solutions, extreme situations (such as heavy rain causing water accumulation in foundation pits) are simulated in the digital twin environment, the feasibility of the construction plan is verified in advance, and the cost of on-site trial and error is reduced; risk propagation is blocked: "vulnerable nodes" are identified through complex networks (for example, the failure of a node will cause delays in 10 subsequent processes), and the monitoring intensity is increased in a targeted manner; dynamic alternative switching: when the risk of the main plan exceeds the standard, the backup path is automatically activated (such as using prefabricated components instead of cast-in-place), to ensure the resilience of the construction period, improve construction stability, ensure the refined management of the project construction, and guarantee the construction progress.

[0016] Optionally, after the dynamic startup step, a resource elastic recovery step is also included; Resource elastic recovery includes: resource marking step, dynamic recovery step and compensation scheduling step; Resource tagging: Add recyclable tags to resources allocated to associated nodes, define their minimum occupancy time and recycling trigger conditions; Dynamic recycling: When the demand for resources on the main critical path suddenly increases, the associated node resources are forcibly recycled based on the label priority, and the potential impact of recycling on the node duration is calculated; Compensation scheduling: If the potential impact exceeds the tolerance threshold, the backup resource pool will be automatically called.

[0017] By adopting the above technical solutions, resource ownership is clarified through label management (such as "tower crane-recyclable-priority 2"), avoiding the waste of "occupying but not using" in traditional scheduling; critical path protection: when there is a sudden demand on the main path (such as the early arrival of concrete and the need for rushing work), the resources of the related nodes are forcibly reclaimed to ensure that the core tasks are not obstructed; loss minimization: the backup resource pool (such as temporary equipment leased by agreement) compensates for the impact of recycling and reduces the delay of the construction period of the related nodes; the dynamic recycling mechanism guarantees the supply of critical path resources, label management clarifies the order of resource recycling, and the backup resource pool responds to sudden demand fluctuations.

[0018] Optionally, the elastic buffering step further includes: Progress tracking: Real-time collection of the construction rate of the highest priority node, and calculation of the deviation rate between the actual rate and the predicted rate; Window expansion and contraction: If the deviation rate continues to exceed the set deviation threshold, the remaining construction period is predicted based on the original construction plan, and the buffer time window is dynamically expanded or compressed; Conflict warning: When the buffer time window overlaps with the time window of the main critical path, a conflict warning is triggered.

[0019] By adopting the above technical solutions, anomalies (such as decreased team efficiency) are identified through deviation rates, and the buffer period is adjusted in a timely manner to avoid chain delays. When the time window is expanded, more resources are automatically released, and when it is compressed, resources are recovered to the shared pool, thereby improving resource turnover. When the buffer window overlaps with the key period of the main path (for example, when both need to share the same transport channel), the path is replanned in advance to reduce on-site coordination time. The time window is expanded and contracted to match the actual construction progress, and overlapping warnings are issued to avoid secondary resource conflicts, taking into account both local nodes and the overall construction period requirements.

[0020] Optionally, it also includes a step of collaborative monitoring of construction progress of associated nodes; Collaborative monitoring of construction progress of associated nodes includes: multi-source data synchronization step, progress visualization mapping step and dynamic model update step; Multi-source data synchronization: The IoT gateway collects equipment operation data in the construction area of the associated nodes in real time and synchronizes it with the 4D progress data in the BIM model; Progress visualization mapping: The actual progress of related nodes is superimposed with the planned progress on the GIS map, and the progress deviation areas are marked in the form of heat maps; Dynamic model update: If the actual progress exceeds the preset deviation range, the BIM model lightweight engine is automatically triggered to update the timeline parameters in the 4D progress model and regenerate the visual view.

[0021] By adopting the above technical solutions, UWB positioning data and equipment sensor data are mapped to the BIM model, achieving millisecond-level synchronization of "physical site-digital model" and improving data real-time performance; the red areas in the heat map automatically focus management's attention, improving efficiency compared to traditional report analysis; the WebGL engine is used to achieve second-level loading of TB-level BIM models, ensuring that the adjusted progress can be viewed in real time on mobile devices.

[0022] Optionally, the multi-objective optimization further includes a space occupancy analysis step, an automatic conflict avoidance step, and a progress compensation calculation step; Space occupancy analysis: The 3D boundaries of the construction area of associated nodes are extracted through the BIM model, and the real-time space occupancy status of equipment / personnel is detected in combination with UWB positioning data; Automatic conflict avoidance: If overlapping operating radii of multiple node devices are detected, an avoidance trajectory is generated through a path planning algorithm and sent to the device controller for execution; Progress compensation calculation: Dynamically adjust the working time allocation of subsequent processes based on the delay duration caused by avoidance.

[0023] By adopting the above technical solution, the BIM model accuracy is combined with UWB positioning to identify equipment collision risks within a range of 0.5m; the optimal detour path is generated to reduce equipment downtime, and the delay caused by avoidance is distributed to subsequent non-critical processes to avoid fluctuations in the overall construction period; three-dimensional boundary detection reduces construction conflicts, the path planning algorithm improves equipment collaboration efficiency, and working time is redistributed to maintain the overall rhythm.

[0024] In a second aspect, the present application provides a system that adopts the following technical solutions: A system includes the following modules: Database building module: used to collect bills of quantities, construction schedules, and real-time construction log data to build a dynamically updated project progress database; Real-time data collection module: The input end is connected to the output end of the database construction module, which is used to collect real-time data on labor distribution, equipment operation status and material consumption at the construction site, and synchronize it to the project progress database; Model building module: The input end is connected to the output end of the real-time data acquisition module to generate a visual construction progress model including the time dimension; Progress deviation judgment module: The input end is connected to the output end of the model building module, and is used to calculate the deviation between the planned progress and the actual construction progress. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Deviation analysis module: The input end is connected to the output end of the progress deviation judgment module to perform deviation analysis and generate a resource allocation priority list based on the deviation analysis results; Resource adjustment module: the input end is connected to the output end of the deviation analysis module, and is used to push adjustment instructions to the construction terminal according to the priority list; Simulation module: The input end is connected to the output end of the resource adjustment module, and is used to iteratively simulate the adjusted construction schedule to verify the effectiveness of the resource allocation plan; Redistribution module: The input end is connected to the output end of the simulation module and is used to perform deviation analysis again when the deviation is not eliminated; Optimization module: The input end is connected to the output end of the reallocation module and is used to store historical progress data, deviation reasons and adjustment records in the database.

[0025] By adopting the above technical solutions, the entire chain from data collection, analysis, and execution is unmanned, reducing manual operation errors. Each module can be upgraded independently (such as replacing random forest with neural network) without affecting the overall system operation. It supports PC-side model editing, mobile terminal progress viewing, and large-screen command center monitoring to meet the needs of multi-role collaboration. In summary, this application includes at least one of the following beneficial technical effects: 1. Resource scheduling is based on data analysis to reduce equipment idleness or shortages, improve multi-role collaboration efficiency, and enable data-driven dynamic decision-making, thereby achieving refined project management. 2. Identify parallel construction opportunities through association networks to shorten critical paths; use Bayesian networks to quantify interference probabilities and avoid risks caused by blind parallel construction; reallocate idle resources to improve utilization. Simultaneously, through resource mobilization and allocation, multi-node joint construction can alleviate delays caused by single-line construction. Establishing node associations based on construction locations facilitates simultaneous multi-line construction, reducing the impact of delays on the current mainline construction schedule, reducing resource waste, and ultimately saving costs. 3. 3D boundary detection reduces construction conflicts, path planning algorithms improve equipment collaboration efficiency, and work time redistribution maintains the overall rhythm. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flow chart of the management method in Example 1 of the present application; Figure 2 This is a position sequence flow chart of the associated node intelligent scheduling step in Example 2 of the present application; Figure 3 This is a detailed flow chart of the steps for intelligent scheduling of associated nodes in Example 2 of the present application; Figure 4 This is a detailed flow chart of the risk hedging step in Example 2 of the present application; Figure 5 This is a detailed flowchart of the priority sorting step in Example 2 of the present application; Figure 6 is a detailed flow chart of the multi-objective optimization steps in Example 2 of the present application; Figure 7 is a detailed flow chart of the elastic buffering step in Example 2 of the present application; Figure 8 This is a detailed flow chart of the resource elastic recovery step in Example 2 of the present application; Figure 9 This is a flowchart of the position sequence of the steps for collaborative monitoring of construction progress of associated nodes in Example 2 of the present application; Figure 10 It is a detailed flowchart of the collaborative monitoring steps of the construction progress of associated nodes in Example 2 of this application. DETAILED DESCRIPTION

[0027] The following combination Figures 1 to 10 This application is described in further detail.

[0028] This embodiment discloses a digital project management method for engineering construction.

[0029] Example 1: Reference Figure 1 ,The digital project management method for engineering construction ,includes the following steps: Database construction: ,automatically extract the bill of quantities based on the BIM model, and ,associate it with the construction schedule and the real-time collected ,construction dynamically updated project progress database; Specifically, BIM model parsing: Use Autodesk Revit API to automatically extract the bill of quantities (such as concrete volume, steel bar tonnage) from the BIM model, and map it to international standard codes (such as Uniformat II) to generate a structured quantity database.

[0030] Schedule association: The construction schedule compiled by Microsoft Project (including WBS task nodes, duration, and logical relationships) is associated with the BIM bill of quantities through task ID to form a three-dimensional mapping table of "task-quantity-time".

[0031] Construction log integration: Use OCR (optical character recognition) technology to parse on-site paper logs, combine them with electronic logs entered through the mobile app, extract key data (such as process completion percentage and quality inspection results), and store them in a MySQL database, synchronizing them once a minute.

[0032] Dynamic update mechanism: Use Apache Kafka to build a real-time data pipeline to push BIM changes, schedule adjustments, and log update events to the database to ensure data consistency.

[0033] Real-time data collection: Real-time collection of labor distribution, equipment operation status and material consumption data at the construction site, and synchronization with the progress database; Specifically, for labor distribution: workers wear UWB positioning tags, whose locations are tracked in real time by base stations, and the number of team members in each area is counted using electronic fence technology. The data is transmitted to the cloud via LoRaWAN; or the location of workers is obtained through drone photography, and the personnel situation is reported in the morning team meeting to obtain the personnel activity areas and personnel information.

[0034] Equipment status monitoring: Vibration sensors (IMU) and oil pressure sensors are installed on tower cranes and pump trucks to collect operating parameters (such as speed and fuel consumption). Abnormal data (such as vibration frequency exceeding the limit) triggers MQTT protocol alarms.

[0035] Material consumption tracking: RFID readers (ImpinjR420) are deployed in the rebar processing area to scan rebar bundles with RFID tags. The consumption progress is calculated based on the theoretical usage in the BIM model. A review instruction is triggered when the error exceeds 5%, and material inbound and outbound information is obtained based on the warehouse registration personnel.

[0036] Data synchronization: All real-time data is pre-processed by the edge computing gateway (Huawei Atlas500), written to the time series database InfluxDB, and bidirectionally synchronized with the main database.

[0037] Model building: Combine the BIM model with the construction plan to conduct 4D progress simulation and generate a visual construction progress model that includes the time dimension; Specifically, BIM models are lightweighted: Autodesk Navisworks is used to convert Revit models into NWD format, removing irrelevant components (such as decorative lines), retaining key components such as structure and electromechanical components, and compressing the model volume.

[0038] Timeline binding: In Synchro4D software, associate WBS task nodes with BIM components, set task start / end times and resource allocation (such as crane usage periods), and generate a 4D model with a timestamp.

[0039] Visual rendering: Import 4D models into the web through the Three.js engine, support dragging the timeline to view the daily plan progress, and model components are gradually colored according to the progress completion rate (such as not started - gray, 50% completed - yellow, 100% completed - green).

[0040] Progress deviation judgment: Compare the deviation between the planned progress in the 4D progress model and the actual construction progress. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Specifically, deviation calculation: Every morning, StoredProcedure is called to calculate the difference between the actual progress of each task (based on sensor data, technician reports, and team work summaries) and the planned progress, using a weighted algorithm: ,in, For the task Actual working hours, For the task Planned working hours, For the task The weight of a task is determined by its position on the critical path; is the total critical path work (the sum of the planned work hours for all critical path tasks), The total number of tasks to be assessed during the monitoring cycle; If the task On the critical path, ,in: For the task Earliest start time to total project duration The proportion of time span; If the task If it is not on the critical path, ,in: The total float time of the task.

[0041] If the weight is adjusted dynamically, and if the task weight needs to dynamically respond to progress changes, a time decay factor can be introduced: , is the decay coefficient (e.g. λ = 0.05 means the weight decays by 5% every day); t is the number of days the task is delayed.

[0042] Threshold warning: Set three levels of thresholds (mild deviation 5%, moderate deviation 10%, severe deviation 15%). When the deviation rate exceeds 5%, the system automatically sends a text message / email to the project manager. When it exceeds 10%, the enterprise WeChat robot is triggered to notify executives.

[0043] Root cause identification: The association analysis algorithm (Apriori) is used to mine association rules between deviation tasks and external factors such as weather (access to the Meteorological Bureau API) and supplier delays.

[0044] Deviation analysis: Based on schedule deviations, deviation analysis is performed. Based on the deviation analysis results, the resource gap of the affected processes is dynamically calculated through an algorithm, and the available equipment, personnel, and material inventory data are matched to generate a resource allocation priority list. The deviation analysis step specifically includes: constructing a process resource demand prediction model using a random forest algorithm, inputting historical schedule data, current resource gaps, and environmental variables, and outputting a resource gap priority score for the affected processes; Specifically, process impact chain analysis: a task dependency network is built based on the graph database Neo4j. After inputting the deviation task ID, the subsequent affected processes are traversed through the breadth-first search (BFS) algorithm to generate an impact chain report.

[0045] Resource demand forecasting: A random forest regression model (Python scikit-learn library) is used to input historical resource consumption data, current resource shortages, and environmental variables (temperature, humidity). The model predicts the resource demand for each process over the next three days and outputs information such as "There is a shortage of 8 rebar workers, and 2 more tower crane shifts are needed."

[0046] Define the priority formula: , where the critical path impact , resource scarcity , supplier response speed , the top 20% of tasks will be placed on the emergency allocation list.

[0047] Specifically, process impact chain analysis (graph theory model) defines the task dependency graph: The facility task network is a directed graph ,in: Represents a set of task nodes, each node corresponds to a construction task; Representing the dependency relationship (edge) between tasks can represent the close-before / close-after relationship, that is, Must be in Launch it after completion.

[0048] Influence Chain Analysis Algorithm: For Deviation Tasks , traverse its successor tasks through breadth-first search (BFS) to generate an impact chain set : ; Influencing chain length calculation: Defining tasks Depth of impact For arrive The number of shortest path edges, the total influence chain length is: .

[0049] Resource demand prediction (random forest regression model), input feature vector: For process i, predict the resource demand for the next T days (e.g. T=3) , input features include: historical consumption data: resource consumption in the past N days ; Current gap amount: gap monitored in real time Environmental variables: temperature t, humidity w (collected in real time by sensors or obtained from weather forecasts or official meteorological data).

[0050] Model expression: ;in: is a random forest regression function, which is integrated by M decision trees, and each tree outputs , the final predicted value is: , is the error term, which obeys the normal distribution .

[0051] For example, if the forecast demand for steel workers is The current available number of people is 42, so the gap is 8 people.

[0052] Priority scoring formula (multi-indicator weighted model), defining indicators: Critical path impact , , critical path tasks Resource scarcity , Supplier response speed , ; Priority scoring formula: ; Let the priority score set of all tasks be , sorted in descending order, select the top 20% of tasks: Urgent Task List = ,in .

[0053] For example, in a foundation pit support project, Task A (critical path) , steel reinforcement gap , , Supplier Response Score ; Calculate the priority score: ; If the score distribution of all tasks is: Highest score , the lowest score , 80% quantile ; then task A enters the emergency allocation list.

[0054] Match available resources with the gap list to generate a resource allocation plan that minimizes scheduling costs; verify the allocation plan with cloud-based material inventory data in real time. If the inventory is insufficient, the supplier collaboration interface is automatically triggered to generate an expedited purchase order.

[0055] Specifically, based on the available resources on site and the list of gaps, the scheduling cost is calculated, taking into account transportation costs, labor adjustment costs, rental costs, fixed activation costs, and temporary labor costs, and the resource allocation plan with the lowest cost is selected. When it comes to consumable resources and there are insufficient resources, a purchase order is sent to the supplier.

[0056] Resource adjustment: push adjustment instructions to the construction terminal according to the priority list, and update the resource occupancy status to the database; Specifically, based on the priority list, the resource scheduling engine (OptaPlanner) is invoked to generate the optimal deployment plan. For example, five workers may be transferred from the non-critical path masonry team to the concrete pouring team; or crane No. 3 may be relocated from the steel structure installation area to the core wall construction area between 9:00 AM and 12:00 PM. Instructions are issued: Adjustment instructions are pushed to the construction team's devices (such as smart wristbands or mobile phones) via the RabbitMQ message queue, and the resource adjustment area is simultaneously highlighted in the BIM model. Status updates: Resource occupancy status is written to the blockchain (Hyperledger Fabric) in real time to prevent multiple parties from accessing the same resource simultaneously.

[0057] Simulation: Iterate the simulation of the adjusted construction schedule to verify the effectiveness of the resource allocation plan. If the deviation is not eliminated, reallocate the steps; Specifically, a construction process simulation model was built using AnyLogic. The adjusted resource allocation data was input and the progress over the next seven days was simulated. Key metrics (such as the number of days of shortened construction period and the number of resource conflicts) were output. 5,000 random samplings were performed on uncertainties (such as equipment failure rates) to calculate the success probability of the adjusted plan. If the probability fell below 90%, the plan was deemed invalid. A visual comparison was performed: the progress curves before and after the adjustment were superimposed on the 4D model. The red curve represented the original plan, and the blue curve represented the adjusted forecast. Deviating areas were automatically marked.

[0058] Reassign: Perform deviation analysis again and follow the deviation analysis steps; Optimization: Store historical progress data, deviation reasons, and adjustment records in the database, and use machine learning models to optimize the schedule generation logic for subsequent projects.

[0059] Specifically, if the simulation shows that the deviation has not been eliminated, the system automatically triggers a secondary analysis, introducing a genetic algorithm (NSGA-II) for multi-objective optimization to find a balance between schedule and cost. Knowledge base construction: Each cause of deviation (e.g., "heavy rain caused earthwork to be suspended") is labeled as a knowledge graph node and associated with the response ("activate backup drainage equipment"). Machine learning optimization: An LSTM neural network (using the TensorFlow framework) is used to train historical data and optimize the schedule generation logic. For example, during the rainy season, buffer time for earthwork projects is automatically increased to reduce the amount of open-air rebar processing.

[0060] This embodiment automatically extracts bills of quantities from the BIM model and links them with construction plans and real-time logs (IoT data) to build a dynamic database, addressing the data silos between the design, planning, and execution phases of traditional project management. For example, changes in rebar usage in the BIM model can be synchronized to the schedule in real time, eliminating manual re-entry errors. Apache Kafka and an edge computing gateway enable second-level data updates (such as crane operating status and worker locations). Compared to traditional daily / weekly reporting models, data latency is reduced from hours to minutes, improving decision-making timeliness. A 4D BIM+timeline model enables spatial and temporal overlay comparison of planned and actual progress. Managers can quickly locate problem areas using heat maps, improving the efficiency of traditional manual inspections. Tiered alerts are triggered based on a deviation rate formula (weighted critical path impact). For example, when the deviation rate exceeds 10%, adjustment plans are automatically pushed to executives' mobile phones, avoiding delays caused by traditional escalating reporting. The Apriori algorithm correlates external factors such as weather and suppliers (for example, the confidence level associated with heavy rain and concrete pour delays) to reduce the risk of human error. The random forest regression model integrates historical consumption and environmental variables (such as humidity affecting welding efficiency) to predict resource gaps in the next three days with a low error rate.

[0061] Based on a priority formula (critical path impact + resource scarcity + supplier response speed), deployment plans are dynamically generated. For example, when concrete is in short supply, the system prioritizes deploying pump trucks to the critical path, while non-critical paths utilize backup plans, reducing equipment idleness.

[0062] A resource-time conflict matrix and path planning algorithm are used to reduce equipment space conflicts. A recyclable resource tagging mechanism ensures resource availability on the primary critical path, and the use of a backup resource pool reduces the risk of construction delays. Pre-rehearsing construction processes in a virtual environment (such as simulating foundation pit collapses caused by heavy rain) allows for early identification of potential risks. In one project, two major safety incidents were successfully avoided, saving approximately 3 million yuan in rework costs. A complex network model quantifies the risk propagation path. For example, if a support structure deviation is determined to potentially affect 10 subsequent processes, the system automatically freezes the associated nodes and activates a backup plan, narrowing the impact to two processes. An LSTM model analyzes historical deviation data (such as records of rainy season delays) to automatically optimize subsequent project schedules and improve the accuracy of new project deadline forecasts.

[0063] Example 2: Reference Figure 2 ,The difference between this embodiment and embodiment 1 is that, after the deviation determination step, it also includes an associated node intelligent scheduling step; The steps of intelligent scheduling of associated nodes include: data extraction step, association degree calculation step, interference evaluation step and dynamic start step; Data extraction: Extract task nodes with logical dependencies from the WBS structure tree of the BIM model and construct a process association network diagram, where nodes represent construction tasks and edges represent the preceding / successive relationships between processes; Specifically, the WBS (Work Breakdown Structure) tree in the BIM model is parsed through the Revit API to extract task node attributes (such as task ID, duration, and logical relationships). The relationship between nodes is extracted based on the process association network diagram, and the network diagram is stored in the Neo4j graph database to support efficient traversal and query.

[0064] Correlation calculation: Analyze the centrality of each node in the correlation network, combine the actual impact frequency in historical construction data, and calculate the global correlation score of the node; Specifically, the global relevance score The calculation model is as follows: ;in, Represents topological centrality, using improved betweenness centrality: ;in represents the total number of shortest paths from node i to j, represents the number of shortest paths passing through node k; Indicates the historical impact frequency, normalized: , weight coefficient , .

[0065] Interference assessment: Real-time monitoring of progress deviations of the currently affected process, inferring the potential interference probability of the affected process on the associated nodes through Bayesian networks, and marking the associated nodes as eligible for parallel startup if the interference probability is lower than a preset threshold. Specifically, Bayesian network modeling: constructing a conditional probability table (CPT), the input variables include: the current deviation task r, the floating time FD of the associated node gl; the historical interference probability , interference probability The calculation formula is: , threshold judgment: if , marking the associated nodes as capable of parallel startup.

[0066] Reference Figure 3 and Figure 4 ,The risk hedging steps include : scenario simulation step, impact diffusion analysis step, and hedging strategy generation step; Scenario simulation: Before starting the associated nodes, the digital twin engine simulates the entire construction process to predict possible quality defects and safety risks. Impact Diffusion Analysis: Build a cascading failure model based on complex networks to quantify the propagation path and probability of the impact of construction anomalies at associated nodes on the overall construction period; Hedging strategy generation: If the simulation results show that the risk diffusion value exceeds the tolerance, the node startup instruction will be frozen and alternative plans will be generated until the risk value drops to a safe threshold.

[0067] Dynamic startup: Push instructions to the construction teams corresponding to the nodes that can be started in parallel, allocate idle labor and equipment to the node in advance, lock the inventory reserves of the required materials, generate parallel construction permits and synchronize them to the supervision system.

[0068] Specifically, scenario simulation: Digital twin engine: Use ANSYS TwinBuilder to build a construction process simulation model, input the construction parameters of the associated nodes (such as the concrete pouring rate), and simulate potential risks (such as the probability of formwork expansion). Quality defect prediction: Use finite element analysis to calculate the stress distribution of key components. If it exceeds the safety threshold (such as concrete compressive strength < 90% of the design value), it is marked as high risk. Impact diffusion analysis: Cascade failure model: Construct a directed weighted network, with the node weight being the task duration and the edge weight being the dependency strength. Define cascade propagation rules: ,in is the working period weight of node j, is the distance from node j to the source node; is the attenuation coefficient (taken as 0.1). If the total impact value is greater than the tolerance value θ, the freeze instruction is triggered.

[0069] Hedging strategy generation: If the simulated risk exceeds the limit, the system automatically generates alternative plans: alternative processes: such as changing cast-in-place stairs to prefabricated installation; resource reallocation: drawing resources from non-critical paths to strengthen key node monitoring.

[0070] Dynamic startup: Command push: Start commands are sent to the construction team's smart terminal via the MQTT protocol, including: task number, required resource list, and time window; digital signatures ensure the authenticity of the commands.

[0071] Material reservation: Call the ERP system API to lock inventory and generate a reservation order (such as "50 tons of Φ20 steel bars reserved"); Parallel construction certificate: Generate an NFT certificate based on the smart contract (Solidity), record parallel construction permissions, and synchronize it to the supervision chain (Hyperledger Fabric).

[0072] Reference Figure 3 and Figure 5 , also includes a priority sorting step, which includes: a conflict detection step, a multi-objective optimization step and an elastic buffer step.

[0073] Conflict detection: When multiple associated nodes meet the conditions for parallel startup, a resource-time conflict matrix is constructed to quantify the duration of each node's occupation of key resources and the spatial overlap area; Multi-objective optimization: Establish a mixed integer programming model with the goal of maximizing the duration reduction rate and minimizing resource conflicts. Input the conflict matrix and current resource distribution data, and output the optimal start sequence. Specifically, the resource-time conflict matrix: define a three-dimensional matrix , where q represents the resource type (e.g., tower crane, pump truck); e represents the time slice (divided by hour); o represents the spatial area (grid code divided by the BIM model), Indicates that resource q occupies area o at time e. Conflict quantification: For nodes h and a, calculate the conflict index : , the priority of the startable nodes can be determined by sorting them from small to large according to the conflict index.

[0074] Mixed integer programming model: The objective function is established with the duration reduction rate as the most important indicator and the minimum resource conflict. The objective function is implemented using a weighted sum and addition method. The weight of the duration reduction rate is greater than 0.5 and less than 0.7, and is greater than the weight of the minimum resource conflict.

[0075] Reference Figure 5 and Figure 6 , the multi-objective optimization also includes a space occupancy analysis step, a conflict automatic avoidance step and a progress compensation calculation step; Space occupancy analysis: The 3D boundaries of the construction area of associated nodes are extracted through the BIM model, and the real-time space occupancy status of equipment / personnel is detected in combination with UWB positioning data; Automatic conflict avoidance: If overlapping operating radii of multiple node devices are detected, an avoidance trajectory is generated through a path planning algorithm and sent to the device controller for execution; Specifically, for space occupancy analysis, BIM model extraction uses a Dynamo script to extract the 3D boundaries of the construction area (IFC format) and converts them into OBBs (Oriented Bounding Boxes) for collision detection. UWB positioning fusion deploys Decawave UWB tags to obtain device coordinates in real time (with an accuracy of ±10cm) and map them to the BIM spatial grid.

[0076] Automatic Conflict Avoidance: Path Planning Algorithm: A time-elastic algorithm is used to generate the device avoidance trajectory: Cost = a1 * path length + a2 * time delay, with parameters a1 = 0.6 and a2 = 0.4. Command Distribution: Trajectory data is sent to the device controller via ROS (Robot Operating System).

[0077] Progress compensation calculation: Dynamically adjust the working time allocation of subsequent processes based on the delay duration caused by avoidance.

[0078] Specifically, when there is a time delay due to avoidance, the delay is Δt, and the subsequent process time adjustment formula is: ,in, is the remaining total float time.

[0079] Elastic buffering: Dynamic buffering time windows are allocated to the highest-priority nodes. If their actual progress lags behind the predicted value, resource pullback instructions are automatically triggered to ensure that the main critical path is not disturbed.

[0080] Reference Figure 5 and Figure 7 , the elastic buffering step further includes: Progress tracking: Real-time collection of the construction rate of the highest priority node, and calculation of the deviation rate between the actual rate and the predicted rate; Window expansion and contraction: If the deviation rate continues to exceed the set deviation threshold, the remaining construction period is predicted based on the original construction plan, and the buffer time window is dynamically expanded or compressed; Conflict warning: When the buffer time window overlaps with the time window of the main critical path, a conflict warning is triggered.

[0081] Specifically, an initial buffer time window is preset for the highest priority task nodes (such as the core process on the critical path), and its length is dynamically adjusted according to the real-time progress to ensure flexible resource allocation and controllable construction period. Initial buffer setting: Based on historical data and expert experience, the initial buffer time is allocated at 10%-15% of the task's planned working hours: , ; The time window range is defined as: .

[0082] Buffer tag binding: Add an "elastic buffer" attribute tag to the task in the BIM model and scheduling system to record the buffer window parameters (start time, end time, and scalable range).

[0083] Progress tracking and deviation rate calculation: Data collection: Sensor deployment: Install GPS / IMU modules on construction equipment (such as tower cranes and pump trucks) to collect operating time and workload (such as concrete pouring volume) in real time; Manual reporting: Enter the number of team members and working hours (such as the completion percentage of steel bar binding) through the mobile app.

[0084] Construction rate calculation: The planned rate is obtained by the planned workload and the planned working hours. The actual rate is obtained by dividing the completed workload by the elapsed time. The deviation rate is obtained by subtracting the difference between the actual rate and the planned rate and dividing it by the planned rate. When the deviation lasts for 2 hours, the buffer adjustment mechanism is triggered; based on the direction (positive / negative) and magnitude of the deviation rate, the buffer time window is dynamically expanded or compressed, and resource conflicts are checked in real time.

[0085] Extension rules (progress delayed): If (Actual rate is lower than planned): , for example, the current buffer is 3 days, , then the new buffer time is 3*1.2=3.6 days.

[0086] Compression rules (progress ahead): If (Actual rate is higher than planned): Calculate the compressible time: , update the buffer window: , retaining a minimum of 50% of the initial buffer to prevent risks caused by excessive compression), dynamically adjust the time window range in the scheduling system and synchronize it with the timeline properties of the BIM model.

[0087] Conflict warning and resource withdrawal: Real-time detection of resource conflicts between the buffer time window and the main critical path, triggering warnings and automatically reclaiming resources to ensure that the critical path is not disturbed.

[0088] Conflict detection algorithm: Time overlap judgment: Let the main critical path time window be , the elastic buffer window is , calculate the overlapping interval: ;like , it is determined to be a conflict.

[0089] Resource withdrawal mechanism: Priority sorting: Determine the withdrawal order based on resource type (tower crane > pump truck > manpower); Forced withdrawal command: Send a stop command to the device controller via the MQTT protocol to release resources to the main critical path; Impact assessment: Calculate the impact of resource withdrawal on buffer tasks: ,like , triggering the call of backup resources.

[0090] Early warning push: Project managers and supervisors are notified through three-level early warning channels: corporate WeChat, SMS, and BIM model pop-up windows. Early warning information includes conflict location (BIM model coordinates), resource type, and recommended measures (such as adjusting the construction sequence).

[0091] Elastic buffer visualization and collaboration: Dynamically render the buffer time window status on the BIM+GIS platform to support multi-role collaborative decision-making; Visual mapping: Use gradient colors to identify the buffer window status in the BIM model: green: sufficient buffer (δ∈[-10%, 10%]); yellow: buffer compression (δ>10%); red: buffer expansion (δ<-10%).

[0092] Project managers can manually adjust the buffer window range by dragging the BIM model timeline. The system will automatically check conflicts and provide feasibility prompts. The supervisor will confirm the adjustment plan through blockchain signatures to ensure traceability of operations.

[0093] Reference Figure 8 In other embodiments, after the dynamic startup step, a resource elastic recovery step is also included; Resource elastic recovery includes: resource marking step, dynamic recovery step and compensation scheduling step.

[0094] Resource tagging: Add a "reclaimable" tag to resources allocated to associated nodes, define their minimum occupancy time and recycling trigger conditions; Dynamic recycling: When the demand for resources on the main critical path suddenly increases, the associated node resources are forcibly recycled based on the label priority, and the potential impact of recycling on the node duration is calculated; Compensation scheduling: If the potential impact exceeds the tolerance threshold, the backup resource pool will be automatically called.

[0095] Specifically, smart tags are added to the resources allocated to associated nodes (non-critical path tasks) to clarify their recyclable attributes and constraints, and to achieve dynamic management of resource ownership; recyclable marking: a Boolean field is added to the resource database, and the minimum occupancy time is set according to the process characteristics (such as a minimum of 24 hours for concrete curing); recycling trigger conditions: the main critical path resource gap rate is >15%; the associated node progress advance rate is >10%.

[0096] Priority coefficient : , (the higher the value, the higher the recycling priority, ranging from 0 to 1); Dynamic reclamation decision and execution: When the resource demand on the primary critical path (CP) increases suddenly, resource reclamation is triggered based on label priority sorting and the impact on associated nodes is evaluated.

[0097] Demand surge detection: Real-time monitoring of CP task resource gaps and calculation of the gap rate, which is equal to the ratio of actual demand minus available capacity to implemented demand. When the gap rate is greater than 15% for three consecutive samplings, the recycling process is triggered.

[0098] Recycling resource selection: Candidate set screening: Retrieve all Boolean fields and resources whose occupied time is greater than the minimum occupied time; Priority sorting: sort by Values are sorted in descending order (such as tower crane >Pump Truck ); Recovery volume calculation: Recovery volume = min (CP gap volume, ∑ candidate resource volume) Forced recycling execution: Send an emergency stop command to the device controller through the Industrial Internet of Things protocol (such as OPC UA); update the resource status to the blockchain (Hyperledger Fabric) to prevent repeated calls.

[0099] Recycling impact assessment and compensation scheduling: Quantify the potential impact of recycling operations on the duration of associated nodes, and automatically call on backup resources for compensation when the threshold is exceeded.

[0100] Potential impact calculation: Construction delay prediction: ; The default resource elasticity coefficient is 0.8.

[0101] Tolerance threshold judgment: , which is considered acceptable; otherwise, compensation scheduling is triggered.

[0102] Backup resource pool call: Resource matching: Select similar resources from the backup pool (such as tower cranes leased under an agreement); Economic feasibility check: Ensure that rental cost < recovery income: income = CP delay loss - rental cost, where delay loss is calculated based on the contract penalty per day; Automatic contract signing: Generate a rental order and pay the advance payment through a smart contract (Solidity).

[0103] Flexible recycling visualization and collaboration: Dynamically display resource flow paths and impact ranges on the digital twin platform, supporting multi-role collaborative decision-making.

[0104] BIM+GIS visualization: Recycled resources appear flashing red in the model; flow paths are marked with orange arrows (from associated nodes to the primary critical path); potential impact values are displayed as floating labels. Supervisors, construction teams, and project managers are assigned different permissions to collaborate. Resource recovery events automatically generate reports, including a list of recovered resources, the potential impact value calculation process, and the compensation scheduling decision chain (algorithm log + manual confirmation records).

[0105] Reference Figure 9 and Figure 10 , in other embodiments, it also includes the step of collaborative monitoring of construction progress of associated nodes; Collaborative monitoring of construction progress of associated nodes includes: multi-source data synchronization step, progress visualization mapping step and dynamic model update step; Multi-source data synchronization: The IoT gateway collects equipment operation data in the construction area of related nodes in real time and synchronizes it with the 4D progress data in the BIM model; Progress visualization mapping: The actual progress of related nodes is superimposed with the planned progress on the GIS map, and the progress deviation areas are marked in the form of heat maps; Dynamic model update: If the actual progress exceeds the preset deviation range, the BIM model lightweight engine is automatically triggered to update the timeline parameters in the 4D progress model and regenerate the visual view.

[0106] The embodiment of the present application also discloses a digital project management system for engineering construction.

[0107] The digital project management system for engineering construction includes the following modules: Database building module: used to collect bills of quantities, construction schedules, and real-time construction log data to build a dynamically updated project progress database; Real-time data collection module: The input end is connected to the output end of the database construction module, which is used to collect real-time data on labor distribution, equipment operation status and material consumption at the construction site, and synchronize it to the project progress database; Model building module: The input end is connected to the output end of the real-time data acquisition module to generate a visual construction progress model including the time dimension; Progress deviation judgment module: The input end is connected to the output end of the model building module, and is used to calculate the deviation between the planned progress and the actual construction progress. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Deviation analysis module: The input end is connected to the output end of the progress deviation judgment module to perform deviation analysis and generate a resource allocation priority list based on the deviation analysis results; Resource adjustment module: the input end is connected to the output end of the deviation analysis module, and is used to push adjustment instructions to the construction terminal according to the priority list; Simulation module: The input end is connected to the output end of the resource adjustment module, and is used to iteratively simulate the adjusted construction schedule to verify the effectiveness of the resource allocation plan; Redistribution module: The input end is connected to the output end of the simulation module and is used to perform deviation analysis again when the deviation is not eliminated; Optimization module: The input end is connected to the output end of the reallocation module and is used to store historical progress data, deviation reasons and adjustment records in the database; The intelligent scheduling module for associated nodes: Its input is connected to the output of the progress deviation judgment module. It is used to analyze progress deviations and associated nodes, calculate whether associated nodes can be started in parallel, and calculate the optimal target based on whether there are conflicts among associated nodes, the duration reduction rate of associated nodes, and the minimum resource conflict. It outputs the start sequence and sets a buffer time window for the highest priority. It also simulates risk hedging and recovers resources based on resource allocation and occupancy. Associated node construction progress collaborative monitoring module: The input end is connected to the output end of the associated node intelligent scheduling module, which is used for multi-party monitoring of the construction progress.

[0108] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A digital project management method for engineering construction, characterized by: The following steps are involved: Database construction: Automatically extract the bill of quantities based on the BIM model and associate it with the construction schedule and real-time collected construction log data to build a dynamically updated project progress database; Real-time data collection: Real-time collection of labor distribution, equipment operation status and material consumption data at the construction site, and synchronization with the progress database; Model building: Combine the BIM model with the construction plan to conduct 4D progress simulation and generate a visual construction progress model that includes the time dimension; Progress deviation judgment: Compare the deviation between the planned progress and the actual construction progress in the visual construction progress model. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Deviation analysis: Deviation analysis is performed based on schedule deviations. Based on the results of the deviation analysis, the resource gap of the affected processes is dynamically calculated through an algorithm, and the available equipment, personnel, and material inventory data are matched to generate a resource allocation priority list; Resource adjustment: push adjustment instructions to the construction terminal according to the priority list, and update the resource occupancy status to the database; Simulation: Iterate the simulation of the adjusted construction schedule to verify the effectiveness of the resource allocation plan. If the deviation is not eliminated, perform the reallocation step; Reassign: Perform deviation analysis again and follow the deviation analysis steps; Optimization: Store historical progress data, deviation reasons, and adjustment records in the database, and use machine learning models to optimize the schedule generation logic for subsequent projects.

2. The digital project management method for engineering construction according to claim 1, characterized in that: The deviation analysis step specifically includes: constructing a process resource demand prediction model using a random forest algorithm, inputting historical progress data, current resource gaps, and environmental variables, and outputting a resource gap priority score for the affected process; Match available resources with the gap list to generate a resource allocation plan that minimizes scheduling costs; The allocation plan is verified with the cloud-based material inventory data in real time. If the inventory is insufficient, the supplier collaboration interface is automatically triggered to generate an expedited purchase order.

3. The digital project management method for engineering construction according to claim 1 or 2, characterized in that: After the progress deviation judgment step, the process also includes an associated node intelligent scheduling step; The steps of intelligent scheduling of associated nodes include: data extraction step, association degree calculation step, interference evaluation step and dynamic start step; Data extraction: Extract task nodes with logical dependencies from the WBS structure tree of the BIM model and construct a process association network diagram, where nodes represent construction tasks and edges represent the preceding / successive relationships between processes; Correlation calculation: Analyze the centrality of each node in the correlation network, combine the actual impact frequency in historical construction data, and calculate the global correlation score of the node; Interference assessment: Real-time monitoring of progress deviations of the currently affected process, inferring the potential interference probability of the affected process on the associated nodes through Bayesian networks, and marking the associated nodes as eligible for parallel startup if the interference probability is lower than a preset threshold. Dynamic startup: Push instructions to the construction teams corresponding to the nodes that can be started in parallel, allocate idle labor and equipment to the node in advance, lock the inventory reserves of the required materials, generate parallel construction permits and synchronize them to the supervision system.

4. The digital project management method for engineering construction according to claim 3 is characterized by: The step of intelligently scheduling associated nodes further includes a step of prioritizing: Prioritization includes: conflict detection step, multi-objective optimization step and elastic buffer step; Conflict detection: When multiple associated nodes meet the conditions for parallel startup, a resource-time conflict matrix is constructed to quantify the duration of each node's occupation of key resources and the spatial overlap area; Multi-objective optimization: Establish a mixed integer programming model with the goal of maximizing the duration reduction rate and minimizing resource conflicts. Input the conflict matrix and current resource distribution data, and output the optimal start sequence. Elastic buffering: Dynamic buffering time windows are allocated to the highest-priority nodes. If their actual progress lags behind the predicted value, resource withdrawal instructions are automatically triggered to reduce interference with the main critical path.

5. The digital project management method for engineering construction according to claim 3 is characterized by: Before the dynamic startup step, the risk hedging step is also included: The risk hedging steps include: scenario simulation step, impact diffusion analysis step and hedging strategy generation step; Scenario simulation: Before starting the associated nodes, the digital twin engine simulates the entire construction process to predict possible quality defects and safety risks. Impact Diffusion Analysis: Build a cascading failure model based on complex networks to quantify the propagation path and probability of the impact of construction anomalies at associated nodes on the overall construction period; Hedging strategy generation: If the simulation results show that the risk diffusion value exceeds the tolerance, the node startup instruction will be frozen and alternative plans will be generated until the risk value drops to a safe threshold.

6. The digital project management method for engineering construction according to claim 3 is characterized by: After the dynamic startup step, a resource elastic recovery step is also included; Resource elastic recovery includes: resource marking step, dynamic recovery step and compensation scheduling step; Resource tagging: Add recyclable tags to resources allocated to associated nodes, define their minimum occupancy time and recycling trigger conditions; Dynamic recycling: When the demand for resources on the main critical path suddenly increases, the associated node resources are forcibly recycled based on the label priority, and the potential impact of recycling on the node duration is calculated; Compensation scheduling: If the potential impact exceeds the tolerance threshold, the backup resource pool will be automatically called.

7. The digital project management method for engineering construction according to claim 4 is characterized by: The elastic buffering step further comprises: Progress tracking: Real-time collection of the construction rate of the highest priority node, and calculation of the deviation rate between the actual rate and the predicted rate; Window expansion and contraction: If the deviation rate continues to exceed the set deviation threshold, the remaining construction period is predicted based on the original construction plan, and the buffer time window is dynamically expanded or compressed; Conflict warning: When the buffer time window overlaps with the time window of the main critical path, a conflict warning is triggered.

8. The digital project management method for engineering construction according to claim 3 is characterized by: It also includes steps for collaborative monitoring of construction progress at associated nodes; Collaborative monitoring of construction progress of associated nodes includes: multi-source data synchronization step, progress visualization mapping step and dynamic model update step; Multi-source data synchronization: The IoT gateway collects equipment operation data in the construction area of the associated nodes in real time and synchronizes it with the 4D progress data in the BIM model; Progress visualization mapping: The actual progress of related nodes is superimposed with the planned progress on the GIS map, and the progress deviation areas are marked in the form of heat maps; Dynamic model update: If the actual progress exceeds the preset deviation range, the BIM model lightweight engine is automatically triggered to update the timeline parameters in the 4D progress model and regenerate the visual view.

9. The digital project management method for engineering construction according to claim 4, characterized in that: The multi-objective optimization further includes a space occupancy analysis step, a conflict automatic avoidance step, and a progress compensation calculation step; Space occupancy analysis: The 3D boundaries of the construction area of associated nodes are extracted through the BIM model, and the real-time space occupancy status of equipment / personnel is detected in combination with UWB positioning data; Automatic conflict avoidance: If overlapping operating radii of multiple node devices are detected, an avoidance trajectory is generated through a path planning algorithm and sent to the device controller for execution; Progress compensation calculation: Dynamically adjust the working time allocation of subsequent processes based on the delay duration caused by avoidance.

10. A system for implementing the engineering construction digital project management method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Database building module: used to collect bills of quantities, construction schedules, and real-time construction log data to build a dynamically updated project progress database; Real-time data collection module: The input end is connected to the output end of the database construction module, which is used to collect real-time data on labor distribution, equipment operation status and material consumption at the construction site, and synchronize it to the project progress database; Model building module: The input end is connected to the output end of the real-time data acquisition module to generate a visual construction progress model including the time dimension; Progress deviation judgment module: The input end is connected to the output end of the model building module, and is used to calculate the deviation between the planned progress and the actual construction progress. When the deviation exceeds the preset threshold, a progress delay warning signal is triggered; Deviation analysis module: The input end is connected to the output end of the progress deviation judgment module to perform deviation analysis and generate a resource allocation priority list based on the deviation analysis results; Resource adjustment module: the input end is connected to the output end of the deviation analysis module, and is used to push adjustment instructions to the construction terminal according to the priority list; Simulation module: The input end is connected to the output end of the resource adjustment module, and is used to iteratively simulate the adjusted construction schedule to verify the effectiveness of the resource allocation plan; Redistribution module: The input end is connected to the output end of the simulation module and is used to perform deviation analysis again when the deviation is not eliminated; Optimization module: The input end is connected to the output end of the reallocation module and is used to store historical progress data, deviation reasons and adjustment records in the database.

Images