Digital twinning progress tracking and risk pre-control method for hydraulic engineering construction

Abstract

The invention discloses a digital twinning progress tracking and risk pre-control method for hydraulic engineering construction, and relates to the field of hydraulic engineering construction, and the method comprises the steps: obtaining mechanical positioning, material entering and environment monitoring data through a multi-source sensing terminal, and generating a time-space aligned construction situation data set; a progress weight template is activated based on the construction stage identifier, and a dynamic weight vector is generated in combination with the environment data; comparing the construction situation data with the reference model to calculate a comprehensive progress deviation index; when the index exceeds the limit, constructing a resource topological relation graph to generate an elastic scheduling instruction; outputting a derivative risk probability through the risk conduction model and generating an optimized scheduling instruction set; triggering a calibration instruction to feed back a correction parameter according to the execution data; by establishing a three-level calibration mechanism linked with the construction progress, the risk change and the environmental disturbance, layered accurate adjustment of data acquisition, environmental factors and model parameters is realized.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy project construction, and specifically to a digital twin progress tracking and risk pre-control method for water conservancy project construction. Background Art

[0002] Water conservancy project construction is a highly complex, systematic undertaking, involving multiple dimensions, including mechanical operation, material transportation, environmental adaptation, and multi-disciplinary collaboration. Project management requires real-time monitoring of construction progress, dynamic resource allocation, and proactive risk management to ensure on-time completion and quality compliance.

[0003] However, due to the complex and changing construction environment (such as differences in geological conditions and climate disturbances), the dense cross-processing (such as the connection between foundation excavation and concrete pouring), and the highly dynamic interaction of resources (fluctuations in the efficiency of machinery, transportation, and manpower), traditional management models face severe challenges. Currently, digital twin technology, as a bidirectional mapping tool between physical entities and virtual models, provides a new technical path for construction management, but its application in the field of water conservancy projects still faces the following bottlenecks:

[0004] Insufficient data fusion capabilities: Multi-source sensor data (machine positioning, material delivery, environmental parameters) are difficult to align effectively due to inconsistent collection frequencies and spatiotemporal benchmarks, leading to distorted progress assessments.

[0005] Lack of dynamic adaptability: The existing progress model uses fixed weight distribution, which cannot reflect the differences in sensitivity of different construction stages (such as cofferdam construction and concrete curing) to environmental disturbances;

[0006] Resource scheduling mismatch: Traditional scheduling systems rely on two-dimensional networks, ignoring the three-dimensional relationship between machine performance parameters (fuel consumption rate, load capacity), path dynamic characteristics (slope, traffic density), and personnel skills;

[0007] Risk transmission blind spots: The existing risk warning mechanism only targets local progress deviations and lacks quantitative analysis of the chain reactions caused by sudden changes in construction intensity in geologically sensitive areas, resulting in delayed warning of secondary disasters.

[0008] Rigid feedback mechanism: The calibration strategy uses a fixed cycle and a unified parameter adjustment mode, which cannot dynamically optimize the data collection frequency and model parameters according to the characteristics of the construction stage and environmental disturbances.

[0009] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0010] In response to the shortcomings of the existing technology, the present invention provides a digital twin progress tracking and risk prevention method for water conservancy project construction.

[0011] In order to achieve the above object, the technical solution of the present invention is as follows:

[0012] In a first aspect, the present invention discloses a digital twin progress tracking and risk pre-control method for water conservancy project construction, comprising the following steps:

[0013] Acquire mechanical positioning data, material delivery data, and environmental monitoring data of each construction unit through a multi-source sensing terminal, and generate a spatiotemporally aligned construction status dataset based on the mechanical positioning data, material delivery data, and environmental monitoring data;

[0014] Obtaining a current construction stage identifier, activating a corresponding progress weight template based on the construction stage identifier, and generating a dynamic weight vector in combination with the environmental monitoring data;

[0015] Comparing the construction status data set with the baseline progress model, and calculating the comprehensive progress deviation index in combination with the dynamic weight vector;

[0016] When the comprehensive progress deviation index exceeds the warning threshold, a resource topology relationship map is constructed, and a performance parameter set associated with the resource topology relationship map is retrieved to generate a flexible scheduling instruction;

[0017] Input the flexible scheduling instructions into the risk transmission model, output the derived risk probability of each construction unit, and generate an optimized scheduling instruction set marking the risk hotspot;

[0018] The actual execution data of the optimized scheduling instruction set is obtained, and whether the third-level calibration instruction is triggered is determined according to the actual execution data. If triggered, the correction parameters are fed back to the perception terminal.

[0019] In a second aspect, the present invention discloses a digital twin progress tracking and risk pre-control system for water conservancy project construction, comprising:

[0020] Data acquisition module: used to obtain the mechanical positioning data, material delivery data and environmental monitoring data of each construction unit through multi-source sensing terminals, and generate a spatiotemporally aligned construction status dataset based on the mechanical positioning data, material delivery data and environmental monitoring data;

[0021] The progress tracking module is used to obtain the current construction stage identifier, activate the corresponding progress weight template based on the construction stage identifier, and generate a dynamic weight vector in combination with the environmental monitoring data; compare the construction status dataset with the baseline progress model, and calculate the comprehensive progress deviation index in combination with the dynamic weight vector; when the comprehensive progress deviation index exceeds the warning threshold, construct a resource topology relationship map, call the associated efficiency parameter set in the resource topology relationship map, and generate a flexible scheduling instruction;

[0022] Risk pre-control module: used to input the flexible scheduling instructions into the risk transmission model, output the derived risk probability of each construction unit, and generate an optimized scheduling instruction set marking the risk hotspot;

[0023] Feedback calibration module: used to obtain the actual execution data of the optimized scheduling instruction set, determine whether to trigger the third-level calibration instruction based on the actual execution data, and if triggered, feed back correction parameters to the perception terminal.

[0024] Compared with the prior art, the present invention has the following beneficial effects:

[0025] 1. By adding synchronized timestamps to machine positioning data, material arrival records, and environmental parameters, and establishing a difference threshold trigger verification mechanism, we can eliminate the temporal misalignment of multi-source perception data, ensure the spatiotemporal consistency of the construction status dataset, and avoid misjudgment of process connections caused by asynchronous data collection.

[0026] 2. Activate differentiated weight templates based on construction stage identifiers (e.g., concrete curing, underground structure construction), and dynamically modify weight vectors based on real-time rainfall intensity, temperature gradient, and curing time factors. This enables precise quantification of the progress deviation index and significantly improves the matching of resource scheduling strategies under complex environmental disturbances.

[0027] 3. Construct a three-dimensional topological map that includes mechanical node efficiency parameters (fuel consumption rate, load capacity), path dynamic characteristics (slope, traffic density), and personnel skill levels. Combined with real-time edge weight calculation, it generates flexible scheduling instructions, reducing the risk of secondary delays caused by improper transportation path selection and improving resource utilization.

[0028] 4. Through the normalized calculation of geological sensitivity coefficient and construction intensity change, the derived risk probability is quantified and the three-dimensional risk hotspot is marked. The optimized scheduling instructions and feedback calibration mechanism are triggered simultaneously to achieve accurate blocking of the risk transmission path, improve the accuracy of secondary geological disaster warning, and shorten the progress recovery time. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0030] Figure 1 This is a block diagram of the overall method of embodiment 1 of the present invention;

[0031] Figure 2 This is a flow chart of a method according to embodiment 1 of the present invention;

[0032] Figure 3 This is an overall system block diagram of the second embodiment of the present invention. DETAILED DESCRIPTION

[0033] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0034] Application Overview: In existing technologies, water conservancy project construction management mostly relies on manual inspections and static progress plans, making it difficult to perceive changes in construction status in real time. Traditional methods use a single data source or discrete environmental sensor data, lacking a multi-dimensional data fusion mechanism aligned in time and space, resulting in delayed identification of progress delay risks. For example, the causal relationship between delayed material delivery and stagnation of mechanical operations is often impossible to accurately trace due to asynchronous data collection. Existing progress assessment models use fixed weight distribution and do not consider the dynamic coupling effect of construction stage characteristics and environmental disturbances, resulting in a mismatch between resource scheduling strategies and actual construction conditions.

[0035] To address these issues, the inventors discovered that asynchronous data collection and static weight assignment are key factors limiting progress tracking accuracy. By analyzing the correlation between construction phase characteristics and environmental parameters, they proposed a dynamic weight adjustment mechanism. To address resource scheduling mismatches, they constructed a resource topology network model to reflect dynamic performance parameters. Furthermore, considering the risk transmission effect, they established a risk hotspot marking and feedback calibration mechanism to achieve closed-loop control.

[0036] Example 1:

[0037] like Figure 1-2 As shown in the figure, a digital twin progress tracking and risk prevention method for water conservancy project construction includes: obtaining mechanical positioning, material delivery and environmental monitoring data through multi-source sensing terminals to generate a time-space aligned construction status data set; activating a progress weight template based on a construction stage identifier and generating a dynamic weight vector in combination with environmental data; comparing the construction status data with the benchmark model to calculate a comprehensive progress deviation index; constructing a resource topology relationship map to generate flexible scheduling instructions when the index exceeds the limit; outputting the derived risk probability through a risk conduction model and generating an optimized scheduling instruction set; triggering calibration instructions based on execution data to feedback correction parameters.

[0038] Among them, the multi-source sensing terminal refers to a data acquisition device that integrates a GPS positioning module, an RFID identification device, and an environmental sensor. Specifically, it can be implemented by combining a Beidou positioning terminal with a temperature and humidity sensor to synchronously acquire mechanical, material, and environmental data. The dynamic weight vector refers to a weight distribution matrix that is dynamically adjusted according to the construction stage and environmental parameters. Specifically, it can be generated by correcting the basic weight template by rainfall intensity and temperature gradient to reflect the differences in environmental sensitivity at different stages. The resource topology relationship map refers to a three-dimensional network model that includes mechanical, transportation, and human nodes and their efficiency parameters. Specifically, a graph database can be used to construct a topological structure with weighted edge connections to optimize resource scheduling paths. The risk transmission model refers to an algorithm that calculates risk probability based on geological sensitivity coefficients and changes in construction intensity. Specifically, a Bayesian network can be used to simulate risk transmission paths to identify potential risk hotspots.

[0039] Specifically, during the foundation excavation phase, multi-source sensing terminals collect real-time data on excavator positioning coordinates and earthmoving vehicle arrivals, simultaneously recording rainfall and surface temperature. When the deviation between the material arrival timestamp and the mechanical operation timestamp exceeds a threshold, a data re-collection mechanism is triggered to ensure spatiotemporal alignment. The system automatically identifies the cofferdam construction phase, retrieves the basic weight template for that phase, and calculates dynamic weights based on real-time rainfall intensity. Comparing the actual construction trajectory with the baseline model reveals that the mechanical operation area of a certain construction unit is lower than expected. The delayed material arrival rate and environmental interference values together contribute to an excessive comprehensive progress deviation index. The system constructs a topological map that includes transportation route slope and vehicle load parameters, generating scheduling instructions for adding backup transport vehicles. After inputting the scheduling plan into the risk model, it predicts the risk of geological disturbance caused by increased transportation pressure in the underground structure construction unit. High-risk areas are marked in the 3D model and detour route instructions are generated. Based on the progress recovery rate and risk entropy changes two hours after the execution of the instructions, a secondary calibration instruction is triggered to adjust the weight of the temperature sensitivity factor.

[0040] Compared with existing technologies, existing methods use a single GPS data source, resulting in a lack of correlation between mechanical operations and material delivery. This solution achieves multi-source data fusion through a spatiotemporal alignment mechanism. Traditional fixed-weight models are unable to adapt to the temperature-sensitive characteristics of concrete during the curing phase. This solution achieves adaptive weight adjustment through a dynamic correction coefficient matrix. Conventional scheduling strategies ignore the impact of path slope on transportation efficiency. This solution constructs a topological map with efficiency parameters to optimize scheduling paths. Existing risk prevention and control methods lack derivative effect analysis. This solution calculates the risk transmission probability through geological sensitivity coefficients and changes in construction intensity.

[0041] Through the above technical solutions, this application solves the problem of inaccurate progress assessment caused by asynchronous multi-source data collection, achieving real-time perception and accurate diagnosis of construction status. The dynamic weight adjustment mechanism enhances the model's adaptability to complex environmental disturbances and avoids resource mismatches caused by fixed weights. The elastic scheduling strategy based on the topological map improves resource utilization efficiency and reduces the chain reaction caused by local progress delays. The risk transmission model and feedback calibration mechanism form a closed-loop control, effectively suppressing the spread of derivative risks.

[0042] This application further proposes that the specific process of spatiotemporal alignment includes adding a first positioning timestamp to the mechanical positioning data and adding a second entry timestamp to the material entry data; when it is detected that the time difference between the first positioning timestamp and the second entry timestamp of the same construction unit exceeds the set threshold, the data verification instruction is automatically triggered and the related data is re-collected.

[0043] Among them, the first positioning timestamp refers to the precise time mark when the mechanical positioning data is collected. Specifically, it can be achieved by synchronous recording with the GPS timing module and the mechanical sensor to ensure the correspondence between the mechanical operation time and the spatial coordinates. The second entry timestamp refers to the time mark when the material transport vehicle arrives at the construction site. Specifically, it can be automatically generated by an RFID scanning device or a floor scale weighing system to record the actual material entry node. The data verification instruction refers to a trigger signal for re-collecting associated data whose timestamp difference exceeds the allowable range. Specifically, it can be logically judged through a preset time window threshold. For example, when the time interval between mechanical operation and material entry exceeds the planned connection time, it is determined that there is a data collection anomaly. Re-collecting associated data refers to the secondary acquisition of mechanical positioning and material entry information within the abnormal time period. Specifically, it can be achieved by activating backup sensors or calling historical cache data.

[0044] Specifically, during the construction of water conservancy projects, there are timing requirements for the connection between mechanical operations and material transportation. By appending a first positioning timestamp to the mechanical positioning data, such as the acquisition time of the excavator's operating trajectory, and appending a second entry timestamp to the material entry data, such as the time when the cement tanker arrived at the unloading area, both are synchronously recorded in the construction status dataset. When the time difference between the first positioning timestamp and the second entry timestamp within the same construction unit exceeds the set threshold, for example, when the machinery has stopped working but the materials have not yet arrived, the system automatically triggers a data verification instruction. During this process, the verification instruction will interrupt the current data stream and initiate the re-collection of related data, such as re-retrieval of the machinery's positioning log within that period or verification of the transport vehicle's driving recorder data. Through this mechanism, the problem of spatiotemporal data misalignment caused by sensor failure or communication delay can be eliminated.

[0045] Compared with existing technologies, traditional methods often use independent collection cycles for machine positioning and material arrival data, lacking a timestamp alignment mechanism. This leads to logical inconsistencies during data analysis, such as when machine operations are completed but materials have not yet arrived. This solution implements temporal consistency verification for multi-source data through timestamp binding and difference threshold judgment. For example, if there is a time difference of more than 30 minutes between the operating time of the vibrating machine and the aggregate arrival record during the concrete pouring phase, the system automatically triggers data review, avoiding progress misjudgments caused by asynchronous data collection.

[0046] Through the above technical solution, this application effectively solves the problem of temporal and spatial misalignment of mechanical operation and material transportation data in water conservancy project construction caused by asynchronous collection. Through timestamp comparison and automatic verification mechanism, it ensures the temporal consistency of multi-source perception data, provides a reliable data basis for subsequent progress deviation calculation, and avoids the risk of false alarms or missed detections caused by data collection anomalies.

[0047] The present application further proposes that the generation process of the dynamic weight vector includes: the construction stage identifier includes but is not limited to the foundation excavation stage, the cofferdam construction stage, the concrete pouring stage, the concrete curing stage, the grouting operation stage, and the underground structure construction stage; the basic weight vector corresponding to the construction stage identifier is retrieved from the pre-stored weight template library; the rainfall intensity and temperature gradient in the current environmental monitoring data are extracted; the dynamic correction coefficient matrix is calculated, wherein the rainfall sensitivity factor and the temperature sensitivity factor are respectively used to adjust the influence of the rainfall intensity and the temperature gradient on the weight; the dynamic weight vector is generated; when the construction stage identifier is the concrete curing stage, the temperature gradient compensation mode is automatically activated, and the curing time factor is added to the dynamic correction coefficient matrix.

[0048] Among them, the basic weight vector refers to the weight distribution benchmark pre-defined according to the progress-sensitive characteristics of different construction stages. It can be generated by a statistical analysis model based on historical construction data, and is used to reflect the baseline impact ratio of mechanical operations, material arrivals and environmental interference at each stage. The rainfall sensitivity factor refers to the adjustment parameter that quantifies the impact of rainfall intensity on the construction progress. It can be implemented by a regression coefficient based on the rainfall intensity-operation efficiency curve, and is used to dynamically adjust the weight ratio of mechanical operations under rainfall conditions. The temperature gradient compensation mode refers to the introduction of a correction mechanism for the difference between the curing time and the benchmark time during the concrete curing stage. It can be implemented by the linear superposition of time differences, and is used to automatically compensate for weight deviations when temperature fluctuations cause the curing time to deviate from the plan.

[0049] Specifically, the method activates the corresponding basic weight vector through the construction stage identifier, and combines the real-time rainfall intensity and temperature gradient data to construct a dynamic correction coefficient matrix. This matrix uses the rainfall sensitivity factor and the temperature sensitivity factor as adjustment parameters, so that the basic weight vector is adjusted in real time according to the environmental disturbance. For example, during the concrete pouring stage, if there is a sudden heavy rainfall, the rainfall sensitivity factor will automatically increase the weight of the environmental disturbance, thereby more accurately reflecting the suppressive effect of rainfall on construction efficiency in the comprehensive progress deviation index. When entering the concrete curing stage, the system automatically superimposes the curing time factor. If the actual curing time exceeds the benchmark value due to temperature changes, the weight is dynamically corrected by adding the time difference term to avoid substandard curing quality or misjudgment of progress due to temperature fluctuations.

[0050] Compared with existing technologies, traditional methods use fixed weight assignments for progress assessment and fail to consider the dynamic coupling effects of construction stage characteristics and environmental disturbances. For example, the concrete curing stage is significantly more sensitive to temperature gradients than the cofferdam construction stage, but existing models are unable to adaptively adjust the temperature impact weight. This solution activates a differentiated correction mechanism through stage identifiers and introduces rainfall sensitivity factors, temperature sensitivity factors, and curing time factors to achieve dynamic reconstruction of the weight vector. This dual adjustment mechanism based on environmental parameters and stage characteristics addresses the problem of traditional methods' insufficient quantification of environmental factors during critical construction stages.

[0051] Through the above technical solution, this application achieves dynamic and precise adjustment of construction progress assessment weights. During critical stages such as concrete curing, the system can automatically compensate for the environmental impact weights of the progress calculation model based on temperature gradient changes and curing time deviations, avoiding curing quality risks or progress misjudgments due to temperature fluctuations. At the same time, the introduction of rainfall sensitivity factors enables the accurate quantification of mechanical operation efficiency in heavy rainfall environments, thereby improving the calculation reliability of the comprehensive progress deviation index under complex working conditions.

[0052] The present application further proposes a method for calculating a comprehensive progress deviation index, including calculating the proportion of the actual operating area of the machinery based on the machinery positioning data in the construction status dataset and the planned trajectory data in the baseline progress model to generate the machinery operation progress, comparing the actual arrival timestamps and planned arrival timestamps in the material arrival data to calculate the proportion of delay time to generate the material arrival progress, accumulating the duration of rainfall intensity exceeding a threshold in the environmental monitoring data to generate the environmental interference duration, normalizing the machinery operation progress, material arrival progress, and environmental interference duration to obtain three normalized indicators respectively, and calculating the comprehensive progress deviation index based on the dynamic weight vector combined with the normalized indicators. The calculation formula is the weighted sum of each component of the dynamic weight vector and the corresponding normalized indicator.

[0053] The actual machine operation area ratio refers to the ratio of the construction area actually covered by the machine to the planned operation area, calculated using positioning data. This can be achieved using geo-fencing technology and polygon area calculation algorithms. It quantifies the impact of machine operation efficiency on the overall progress. The material delay ratio refers to the ratio of the absolute difference between the actual material arrival time and the planned arrival time to the total duration of the construction phase. This can be achieved using a timestamp comparison and difference calculation module, reflecting the degree of disruption to the construction process caused by material supply delays. The environmental interference duration refers to the cumulative duration of rainfall intensity exceeding a set threshold. This is achieved using a time series accumulation statistical method based on rainfall sensor data. It characterizes the hindering effect of extreme weather on construction activities. Normalization maps progress indicators of different dimensions to a unified numerical range. This can be achieved using range normalization or z-score normalization algorithms to eliminate the impact of magnitude differences between indicators on the comprehensive evaluation. The dynamic weight vector refers to a weight distribution parameter that is dynamically adjusted based on the characteristics of the construction phase and environmental parameters. It is generated by multiplying a preset template with a real-time correction coefficient matrix, enabling differentiated evaluation of various progress-influencing factors under different construction conditions.

[0054] Specifically, by comparing mechanical positioning data with the benchmark model, a mechanical operation progress index is generated to reflect the degree of deviation between equipment operating efficiency and the planned trajectory; by using the delayed statistics of material arrival timestamps, a material arrival progress index is generated to capture the impact of supply chain fluctuations on the construction rhythm; and by using the cumulative duration of excessive rainfall in environmental monitoring data to generate an environmental interference index to quantify progress losses caused by uncontrollable natural factors. The three types of indicators are normalized to eliminate dimensional differences, and then weighted summed according to the weight values of the corresponding components in the dynamic weight vector to obtain a progress deviation index that comprehensively reflects multi-dimensional influencing factors. This calculation method integrates real-time data from three dimensions: mechanical operation, material supply, and environmental interference. Combined with a dynamically adjusted weight distribution mechanism, it can adapt to changes in core contradictions in different construction stages. For example, during the foundation excavation stage, the focus is on monitoring mechanical operation efficiency, while during the concrete curing stage, the weight of environmental interference factors is increased.

[0055] Compared with existing technologies, traditional methods calculate a single progress deviation indicator based solely on machine positioning data, failing to account for the combined impact of material supply delays and environmental disturbances. Furthermore, the fixed weight distribution model fails to reflect the core contradictions at different construction stages. This solution achieves a comprehensive quantitative assessment of progress deviations through multi-dimensional data fusion and dynamic weight adjustment, enabling more accurate identification of the dominant factors contributing to progress delays. For example, during the cofferdam construction phase, which experiences frequent rainfall, the system automatically increases the weight of the environmental disturbance indicator, avoiding overall progress risks caused by ignoring weather impacts while only meeting the requirements for a single mechanical operation indicator.

[0056] Through the above technical solution, this application can solve the problems of single data dimension and rigid weight distribution of traditional progress monitoring methods, and realize multi-factor collaborative evaluation of construction progress deviation. Through the combination of dynamic weights and normalization processing, the key bottleneck factors in different construction scenarios can be effectively identified, and an accurate quantitative basis can be provided for resource scheduling decisions to avoid progress misjudgments caused by improper weight distribution. For example, when material delays and environmental interference occur at the same time, the system automatically balances the influence weights of the two through dynamic weights, accurately reflects the true impact of complex factors on the overall progress, and prevents decision-making deviations caused by overload of a single indicator.

[0057] This application further proposes that the construction of a resource topology relationship map includes establishing a three-dimensional topological network of mechanical nodes, transportation nodes, and human nodes, and associating an efficiency parameter set for each node. The efficiency parameter set includes mechanical fuel consumption rate, transport vehicle load weight, and personnel skill level. Weighted connection edges are established between adjacent nodes, and the edge weights include path slope and traffic density parameters.

[0058] Mechanical nodes are information carriers of the spatial location and status of various types of construction machinery on a construction site. These nodes can be implemented using a combination of Beidou positioning devices and equipment status sensors. For example, excavator nodes upload coordinates and fuel consumption data in real time via onboard terminals. Transport nodes are key location markers along material transport corridors. These nodes can be implemented using RFID landmarks linked to traffic flow detectors, such as smart road signs deployed at construction road intersections. Human nodes are dynamic management units for areas where workers gather. These nodes can be implemented using Bluetooth beacons and personnel positioning bracelets, such as electronic fences installed in rebar processing areas. Path slope refers to the rate of change in terrain elevation between connected nodes. This can be calculated using LiDAR scanning combined with a digital elevation model. Traffic density refers to the number of transport vehicles passing through a connecting edge per unit time. This can be achieved by using video analysis algorithms to calculate lane occupancy.

[0059] Specifically, during the concrete pouring phase, when the edge weight between the concrete mixer truck node and the pumping equipment node changes, the system will automatically update the transport path selection strategy. For example, if the slope value of a transport path increases to a preset threshold due to rainfall, the system will recalculate the edge weight and trigger the path optimization instruction. At the same time, the load parameters of the transport node and the fuel consumption rate parameters of the mechanical node are correlated and analyzed. When the load of the transport vehicle reaches the upper limit, the spare mechanical equipment node with a higher fuel consumption rate is automatically matched to maintain transportation efficiency. In the three-dimensional topology network, the skill level parameters of the human node are dynamically associated with the location information of the mechanical node. When it is detected that the mechanical operation in a certain area is stagnant, the personnel node with the corresponding operating qualifications can be quickly retrieved for scheduling.

[0060] Compared with existing technologies, traditional resource scheduling systems often use two-dimensional planar networks and only consider a single performance parameter, such as allocating resources based solely on transport distance. However, the three-dimensional topological network constructed by this solution integrates multi-dimensional performance parameters such as mechanical status, transport capacity, and personnel skills, and combines them with dynamically changing path slope and traffic density edge weights to more accurately reflect the three-dimensional spatial relationships and resource interaction status of the construction site. For example, during the cofferdam construction phase, traditional methods may ignore the impact of real-time slope changes on vehicle loads. However, this solution can avoid transport interruptions caused by overloaded vehicles by dynamically updating weighted connecting edges.

[0061] Through the above technical solutions, this application can achieve multi-dimensional matching and dynamic optimization of construction resource scheduling, and effectively solve the scheduling inaccuracy problem caused by the traditional method ignoring the correlation between node efficiency parameters and the dynamic characteristics of the path. For example, in the underground structure construction scenario, by associating the skill level of the human node with the spatial position of the mechanical node, a construction team with underground operation qualifications can be quickly formed to avoid progress delays caused by mismatch of personnel skills. At the same time, the dynamic weight calculation of path slope and traffic density provides real-time road condition perception capabilities for resource scheduling, significantly reducing the risk of secondary delays caused by improper transportation path selection.

[0062] The present application further proposes that the process of generating flexible scheduling instructions includes: marking the construction unit whose comprehensive progress deviation index exceeds the warning threshold as a progress-lagging unit; obtaining the node identifier of the progress-lagging unit, and searching the resource topology relationship map for candidate nodes whose weighted sum of the path slope influence factor and the traffic density influence factor is less than the comprehensive scheduling cost threshold; and generating an instruction queue containing the candidate node identifier, the scheduling quantity and the optimal path.

[0063] Among them, the progress lagging unit refers to the construction unit whose comprehensive progress deviation exceeds the allowable range due to mechanical operation progress, material arrival progress or environmental interference factors. Specifically, it can be dynamically marked by using a threshold comparison algorithm combined with the real-time calculation results of the construction status data set. Its function is to accurately locate the target area that requires resource scheduling. Candidate nodes refer to mechanical, transportation or manpower nodes with available resources and controllable scheduling costs in the resource topology relationship map. Specifically, they can be obtained by screening through a linear combination of path slope influencing factors and traffic density influencing factors. Its function is to ensure the economy and feasibility of the resource scheduling plan. The optimal path refers to the route with the highest transportation efficiency calculated based on the edge weights of weighted connecting edges in the three-dimensional topological network. Specifically, it can be dynamically planned using the Dijkstra algorithm combined with real-time updated traffic density parameters. Its function is to reduce the scheduling time of transportation nodes.

[0064] Specifically, when a delayed unit is marked, the system locates the corresponding machinery, transportation, or manpower node in the resource topology relationship map based on the node identifier. The system extracts the candidate node's fuel consumption rate, load capacity, and skill level data through a preset performance parameter set, and calculates the comprehensive scheduling cost of the candidate node by combining the real-time parameters of path slope and traffic density. After screening the candidate nodes that meet the cost constraints, the system generates an instruction queue containing the node number, the number of dispatched equipment, and the real-time optimized path. For example, it can output the operation instruction "Dispatch two excavators from node A3 and transport them via path P5." This process achieves flexible allocation of construction resources by dynamically matching resource demand with node performance.

[0065] Compared to existing methods, which rely on static topological networks for resource scheduling and ignore the dynamic impact of changing path slopes on transportation efficiency, this solution introduces real-time traffic density parameters for weighted links, combined with a dynamic screening mechanism for candidate nodes, enabling scheduling instructions to adapt to changing traffic conditions at the construction site. For example, if a transportation route's slope increases due to rainfall, the system automatically adjusts the optimal path selection strategy to avoid the risk of vehicle delays.

[0066] Through the above technical solution, this application solves the problem of mismatch between traditional scheduling instructions and dynamic construction conditions, and realizes the coupled optimization of resource scheduling schemes and real-time environmental parameters. By screening candidate nodes that meet comprehensive cost constraints and planning the optimal path, the idling rate of mechanical equipment and the waiting time of transportation nodes are effectively reduced, thereby improving resource scheduling efficiency. At the same time, the dynamic calculation mechanism based on the three-dimensional topological network can avoid scheduling failures caused by local path congestion, ensuring the reliability of construction progress recovery.

[0067] The present application further proposes a method for calculating the derivative risk probability, including parsing the flexible scheduling instructions, determining the affected set of associated construction units, and based on the geological sensitivity coefficient and construction intensity change of each construction unit in the associated construction unit set, calculating the derivative risk probability equal to the geological sensitivity coefficient multiplied by the ratio of the construction intensity change to the maximum allowable intensity change.

[0068] The geological sensitivity coefficient is a quantitative parameter that characterizes the sensitivity of the geological conditions of a construction unit to construction disturbances. It can be calculated through a comprehensive calculation of rock mass stability indicators and groundwater permeability coefficients from geological exploration data, and is used to reflect the responsiveness of the geological environment to changes in construction intensity. The change in construction intensity refers to the difference in resource input by a construction unit before and after the execution of a scheduling instruction. It can be calculated by weighted summation of the change in mechanical operation duration, the change in material consumption rate, and the fluctuation in manpower input. It is used to quantify the extent to which scheduling instructions affect the operating status of a construction unit. The maximum allowable intensity change refers to the maximum construction intensity adjustment threshold allowed for a construction unit under safe operating conditions. It can be determined based on the mechanical load limit, material supply rate limit, and manpower allocation safety standards in the construction process specifications, and is used to constrain the reasonable range of risk probability calculation results.

[0069] Specifically, after a flexible scheduling instruction is triggered, the system first analyzes the instruction to identify the set of associated construction units affected by the scheduling operation. For each associated construction unit, the system obtains the geological sensitivity coefficient pre-annotated in the geological exploration report. This coefficient integrates geological parameters such as the degree of rock fracture development, soil shear strength, and groundwater activity. Mechanical operation sensor data and material inventory data from the construction unit are simultaneously collected to calculate the change in construction intensity before and after the scheduling operation. The geological sensitivity coefficient and the change in construction intensity are substituted into a preset risk probability formula and normalized against the maximum allowable intensity change specified in the process specification to generate a derived risk probability value for each unit. When the probability value exceeds a set threshold, the risk hotspot marking process is automatically triggered. For example, if the geological sensitivity coefficient of an underground structure construction unit reaches 0.8 and the scheduling instruction results in a 30% increase in the concrete pouring rate, the system calculates whether the change in construction intensity for that unit exceeds its maximum allowable value and generates a risk warning signal accordingly.

[0070] Compared to existing technologies, traditional approaches assess risk based solely on progress deviations within a single construction unit, failing to consider the impact of scheduling instructions on the geological conditions of associated units and the transmission effects of intensity changes. Existing technologies often store geological parameters and construction intensity data independently, lacking a dynamic correlation analysis mechanism. This approach, by establishing a quantitative correlation model between geological sensitivity coefficients and construction intensity changes, accurately captures chain risk transmission pathways, addressing the issue of delayed identification of derivative risks caused by local scheduling in traditional approaches.

[0071] Through the above technical solution, this application can assess the geological sensitivity and construction intensity trends of related construction units in real time during the resource scheduling process, and identify risk areas exceeding safety thresholds in advance through normalized calculations. This method effectively avoids the risk of secondary geological disasters caused by the execution of scheduling instructions, provides a quantitative decision-making basis for optimizing scheduling instructions, and significantly reduces the probability of chain risk transmission in water conservancy project construction.

[0072] This application further proposes that when the P value of any construction unit exceeds a set threshold, the geographical boundary of the construction unit is highlighted in the three-dimensional construction model; and an optimized scheduling instruction set is simultaneously generated, specifically including supplementary scheduling instructions for alternative route coordinates and emergency resource codes, and path update instructions are sent to the associated transport vehicle terminal.

[0073] A P-value exceeding a set threshold indicates that the derived risk probability calculated by the risk transmission model has reached the preset risk tolerance threshold. This is achieved by using the normalized ratio of the geological sensitivity coefficient to the change in construction intensity, which quantifies the abnormal risk level of a local construction unit. This feature, through a dynamic threshold determination mechanism, accurately identifies high-risk areas, avoiding the misjudgments or omissions caused by traditional methods due to fixed thresholds.

[0074] 3D construction model highlighting visualizes the geographic locations of risk hotspots onto the project's digital twin, using boundary shading or transparency adjustment. This feature addresses the ambiguity in risk location found in existing technologies, enhancing intuitive risk identification through spatial mapping.

[0075] The optimized scheduling instruction set includes backup route coordinates and path update instructions, generated using a real-time path planning algorithm combined with topological network parameters. This feature addresses resource scheduling mismatches by ensuring the accessibility and timely scheduling of emergency resources through multi-dimensional instruction coordination.

[0076] Specifically, when the derived risk probability output by the risk transmission model exceeds a threshold, the system first marks the geographic boundary of the corresponding construction unit in the three-dimensional model, for example, by highlighting it in red or flashing it. At the same time, based on real-time parameters such as path slope and traffic density in the resource topology relationship map, a supplementary dispatch instruction containing the coordinates of the backup transport route is generated, and a route update data packet is sent to the affected vehicles. For example, during the construction phase of an underground structure, if surrounding rock deformation causes a path to be blocked, the system can automatically push a detour route and backup grouting equipment code, realizing spatial linkage between dispatch instructions and risk hotspots.

[0077] Compared to existing technologies, traditional methods typically rely on text alerts or two-dimensional graphic markers, failing to pinpoint risk hotspots in three dimensions. Furthermore, dispatch instructions lack dynamic linkage to geographic information. For example, when a material delay is detected, existing systems only issue a general alert without simultaneously generating a route adjustment plan with specific coordinates. This solution, through 3D model visualization and spatialized instructions, effectively addresses the issues of unclear risk transmission paths and unfocused dispatch measures.

[0078] Through the above-mentioned technical solution, this application achieves precise spatial positioning of risk hotspots and dynamic adaptation of dispatch instructions. In the event of sudden heavy rainfall during the concrete pouring phase, the system can quickly mark the construction area affected by waterlogging, simultaneously push optimized routes to transport vehicles that avoid low-lying sections of road, and deploy pumping equipment to designated coordinates. This mechanism allows the derivative risks caused by local schedule deviations to be quickly suppressed, avoiding the cascading delays caused by traditional methods due to the ambiguous risk transmission paths.

[0079] This application further proposes a feedback correction mechanism that specifically includes the following steps: based on the actual execution data of the optimized scheduling instruction set, synchronously calculate the progress recovery rate δ and the resource scheduling derived risk reduction rate ρ; trigger a three-level calibration instruction when the progress recovery deviation and the risk reduction deviation exceed the tolerance threshold at the same time; execute a hierarchical calibration process including data re-collection, dynamic adjustment of environmental factors and model parameter update, and feed back the calibrated parameters to the perception terminal to adjust the data sampling frequency.

[0080] Among them, the progress recovery rate δ refers to the rate of change of the comprehensive progress deviation index per unit time. Specifically, it can be achieved by calculating the time-varying difference between the current deviation value and the initial deviation value, and is used to quantify the actual recovery effect of the optimized scheduling instructions on the construction progress. The resource scheduling derivative risk reduction rate ρ refers to the relative decrease in the risk entropy value before and after resource scheduling. Specifically, it can be achieved by calculating the change in risk probability output by the risk transmission model, and is used to evaluate the actual suppression effect of scheduling measures on derivative risks. The three-level calibration instruction refers to a progressive calibration trigger mechanism from the data acquisition layer, environmental analysis layer to the model correction layer. Specifically, it can be achieved by logical judgment linked to the difference threshold judgment and the stage identifier to ensure that the calibration measures match the working conditions characteristics of the construction stage. The sampling frequency adjustment formula refers to the calculation rule for dynamically adjusting the data collection period based on the progress recovery deviation rate. Specifically, it can be achieved by weighted calculation of the initial frequency and the deviation rate, and is used to achieve adaptive optimization of the data collection intensity of the perception terminal.

[0081] Specifically, this technical solution establishes dual evaluation metrics, encompassing both progress recovery rate and risk reduction rate, by monitoring the effectiveness of optimized scheduling instructions in real time. When measured metrics deviate significantly from model predictions, a three-level calibration process is initiated. The first level of calibration rechecks timestamp continuity in the base data layer. For example, this triggers additional data collection when mechanical positioning data is interrupted during underground structure construction. The second level of calibration dynamically adjusts weighting factors based on environmental parameter trends. For example, it automatically increases the rainfall sensitivity factor α when the rainfall intensity rate exceeds 5 mm per hour. The third level of calibration updates baseline model parameters based on construction phase characteristics. For example, if abnormal surrounding rock deformation data is detected during cofferdam construction, the deformation monitoring weight ratio is increased from a baseline value of 0.2 to 0.35. The calibrated dynamic weight vectors and model parameters automatically adjust the sensing terminal's sampling frequency through a closed-loop feedback mechanism. For example, if the initial sampling frequency is once per minute, the frequency is increased to 1.2 per minute when the progress recovery deviation rate reaches 20%.

[0082] Compared with existing technologies, existing construction management systems typically rely on fixed calibration cycles and single parameter adjustment mechanisms, such as monthly manual sensor calibration or unified adjustment of monitoring frequencies for all construction units. This solution, by establishing a three-level calibration mechanism linked to construction progress, risk changes, and environmental disturbances, enables hierarchical and precise adjustment of data collection, environmental factors, and model parameters. Compared to traditional methods that misjudge curing durations due to insufficient temperature monitoring frequency during the concrete curing phase, this solution automatically triggers a secondary calibration based on temperature gradient changes and increases monitoring frequency, effectively avoiding curing quality defects.

[0083] Through the above technical solutions, this application solves the problems of model inaccuracy and parameter drift caused by the lack of a closed-loop feedback mechanism in traditional construction management systems. When encountering sudden heavy rainfall during the concrete pouring stage, the system can accurately capture the correlation between material transportation delays and mechanical operation stagnation by dynamically improving the rainfall sensitivity factor and sampling frequency, shortening the scheduling instruction generation time compared to the original fixed weight model. At the same time, in underground structure construction, by activating the weight ratio update mechanism of surrounding rock deformation data, the accuracy of geological risk warning is improved, significantly reducing the risk of support structure instability caused by model parameter lag.

[0084] Example 2:

[0085] like Figure 3 As shown in the figure, the digital twin progress tracking and risk pre-control system for water conservancy project construction includes:

[0086] Data acquisition module: used to obtain the mechanical positioning data, material delivery data and environmental monitoring data of each construction unit through multi-source sensing terminals, and generate a spatiotemporally aligned construction status dataset based on the mechanical positioning data, material delivery data and environmental monitoring data;

[0087] The progress tracking module is used to obtain the current construction stage identifier, activate the corresponding progress weight template based on the construction stage identifier, and generate a dynamic weight vector in combination with the environmental monitoring data; compare the construction status dataset with the baseline progress model, and calculate the comprehensive progress deviation index in combination with the dynamic weight vector; when the comprehensive progress deviation index exceeds the warning threshold, construct a resource topology relationship map, call the associated efficiency parameter set in the resource topology relationship map, and generate a flexible scheduling instruction;

[0088] Risk pre-control module: used to input the flexible scheduling instructions into the risk transmission model, output the derived risk probability of each construction unit, and generate an optimized scheduling instruction set marking the risk hotspot;

[0089] Feedback calibration module: used to obtain the actual execution data of the optimized scheduling instruction set, determine whether to trigger the third-level calibration instruction based on the actual execution data, and if triggered, feed back correction parameters to the perception terminal.

[0090] The above content is merely an example and explanation of the structure of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the scope of protection of the present invention.

[0091] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0092] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A digital twin progress tracking and risk pre-control method for water conservancy project construction, characterized by: The following steps are involved: Acquire mechanical positioning data, material delivery data, and environmental monitoring data of each construction unit through a multi-source sensing terminal, and generate a spatiotemporally aligned construction status dataset based on the mechanical positioning data, material delivery data, and environmental monitoring data; Obtaining a current construction stage identifier, matching a corresponding progress weight template based on the construction stage identifier, and combining the environmental monitoring data to generate a dynamic weight vector; Comparing the construction status data set with the baseline progress model, and calculating the comprehensive progress deviation index in combination with the dynamic weight vector; When the comprehensive progress deviation index exceeds the warning threshold, a resource topology relationship map is constructed, a performance parameter set associated with the resource topology relationship map is retrieved, and a flexible scheduling instruction is generated; Input the flexible scheduling instructions into the risk transmission model, output the derived risk probability of each construction unit, judge the derived risk probability, and generate an optimized scheduling instruction set marking the risk hot zone; The actual execution data of the optimized scheduling instruction set is obtained, and whether the third-level calibration instruction is triggered is determined according to the actual execution data. If triggered, the correction parameters are fed back to the perception terminal.

2. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 1 is characterized in that: The specific process of the spatiotemporal alignment includes: A first positioning timestamp is added to the mechanical positioning data, and a second entry timestamp is added to the material entry data; when it is detected that the time difference between the first positioning timestamp and the second entry timestamp of the same construction unit exceeds the set threshold, the data verification instruction is automatically triggered and the related data is re-collected.

3. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 2 is characterized in that: The generation process of the dynamic weight vector includes: The construction stage identifier includes but is not limited to foundation excavation stage, cofferdam construction stage, concrete pouring stage, concrete curing stage, grouting operation stage, and underground structure construction stage; Retrieve the basic weight vector W_base corresponding to the construction stage identifier from the pre-stored weight template library; Extract rainfall intensity μ and temperature gradient ΔT from current environmental monitoring data; Calculate the dynamic correction coefficient matrix C = 1 + αμ + β |ΔT-T0|, where α is the rainfall sensitivity factor, β is the temperature sensitivity factor, and T0 is the standard temperature gradient; Generate a dynamic weight vector W_dynamic = W_base × C; When the construction phase identifier is the concrete curing phase, the temperature gradient compensation mode is automatically activated: Add the curing time factor γ to the dynamic correction coefficient matrix, so that C = 1 + αμ + β | ΔT - T0 | + γ (t - t0), where t is the actual curing time and t0 is the benchmark curing time.

4. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 3 is characterized in that: The calculation method of the comprehensive progress deviation index includes: Based on the machine positioning data in the construction status dataset and the planned trajectory data in the baseline progress model, the actual machine operation area ratio is calculated to generate the machine operation progress; Compare the actual arrival timestamp with the planned arrival timestamp in the material arrival data, calculate the delay ratio, and generate the material arrival progress; The duration of rainfall intensity exceeding the threshold γ in the accumulated environmental monitoring data is used to generate the environmental disturbance duration; Normalize the mechanical operation progress, material arrival progress, and environmental interference duration to obtain x1, x2, and x3 respectively; Calculate the comprehensive progress deviation index based on the dynamic weight vector W_dynamic=[w1,w2,w3]; The calculation formula is: D=w1x1+w2x2+w3x3.

5. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 4 is characterized in that: The construction of the resource topology relationship map includes: Establish a three-dimensional topological network of mechanical nodes, transportation nodes, and human nodes; Associating each node with a performance parameter set, the performance parameter set including mechanical fuel consumption rate, transport vehicle load, and personnel skill level; Establish weighted connecting edges between adjacent nodes, and the edge weights include path slope and traffic density parameters.

6. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 5 is characterized in that: The process of generating the flexible scheduling instruction includes: The construction units whose comprehensive progress deviation index exceeds the warning threshold are marked as progress lagging units; Obtain the node identifier of the delayed unit and search for candidate nodes that satisfy w1η1+w2η2<θ in the resource topology relationship graph, where η1 is the path slope influencing factor, η2 is the traffic density influencing factor, and θ is the comprehensive scheduling cost threshold; Generate an instruction queue containing candidate node identifiers, scheduling quantities, and optimal paths.

7. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 6 is characterized in that: The method for calculating the derivative risk probability includes: Parse the flexible scheduling instructions and determine the set of related construction units affected; Based on the geological sensitivity coefficient σ and construction intensity variation ΔI of each construction unit in the associated construction unit set, the derived risk probability P = σ × ΔI / Imax is calculated, where Imax is the maximum allowable intensity variation.

8. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 7 is characterized in that: The method for marking the risk hot zone includes: When the P value of any construction unit exceeds the set threshold, the geographical boundary of the construction unit is highlighted in the 3D construction model; An optimized scheduling instruction set is generated simultaneously, specifically including supplementary scheduling instructions for backup route coordinates and emergency resource codes, and path update instructions are sent to associated transport vehicle terminals.

9. The digital twin progress tracking and risk pre-control method for water conservancy project construction according to claim 8 is characterized in that: The feedback correction mechanism specifically includes: According to the actual execution data of the optimized scheduling instruction set, the progress recovery rate δ and the resource scheduling derived risk reduction rate ρ are calculated synchronously, where: δ = (D_current - D_initial) / Δt, where D_current is the comprehensive progress deviation index at the current moment, D_initial is the comprehensive progress deviation index before the instruction is executed, and Δt is the execution time interval; ρ = (E_before - E_after) / E_before × 100%, E_before is the average risk entropy value of each construction unit before scheduling, and E_after is the average risk entropy value after scheduling; When |δ-δ_simulation|>Δδ and |ρ_simulation-ρ|>Δρ are satisfied at the same time, the third-level calibration instruction is triggered, where: δ_ is the progress recovery rate predicted by the risk transmission model, and Δδ is the progress recovery tolerance threshold; ρ_ is the risk reduction rate predicted by the risk transmission model, and Δρ is the risk reduction tolerance threshold; Level 1 calibration: re-collect the mechanical positioning data and material delivery data of the delayed unit and verify the continuity of the timestamp; Secondary calibration: Calculate the environmental trend change rate ξ = Δμ / Δt + Δ(ΔT) / Δt based on the current environmental monitoring data, where Δμ is the change in rainfall intensity and Δ(ΔT) is the change in temperature gradient, and dynamically adjust the rainfall sensitivity factor α and temperature sensitivity factor β in the dynamic weight vector; Level 3 calibration: activating the corresponding model correction mode according to the construction stage identifier. When the identifier is the underground structure construction stage, the weight ratio κ of the surrounding rock deformation monitoring data is increased, and the geological parameter set in the benchmark progress model is updated. Feedback the calibrated dynamic weight vector W'_dynamic and the updated benchmark progress model parameters to the multi-source sensing terminal to automatically adjust the sampling frequency of the machine positioning data; The sampling frequency adjustment formula is: f'=f×(1+|δ-δ_analog| / δ_analog), where f is the initial sampling frequency.

10. The digital twin progress tracking and risk pre-control system for water conservancy project construction is characterized by , using the digital twin progress tracking and risk pre-control method for water conservancy project construction as described in any one of claims 1 to 9, including: Data acquisition module: used to obtain the mechanical positioning data, material delivery data and environmental monitoring data of each construction unit through multi-source sensing terminals, and generate a spatiotemporally aligned construction status dataset based on the mechanical positioning data, material delivery data and environmental monitoring data; The progress tracking module is used to obtain the current construction stage identifier, activate the corresponding progress weight template based on the construction stage identifier, and generate a dynamic weight vector in combination with the environmental monitoring data; compare the construction status dataset with the baseline progress model, and calculate the comprehensive progress deviation index in combination with the dynamic weight vector; when the comprehensive progress deviation index exceeds the warning threshold, construct a resource topology relationship map, call the associated efficiency parameter set in the resource topology relationship map, and generate a flexible scheduling instruction; Risk pre-control module: used to input the flexible scheduling instructions into the risk transmission model, output the derived risk probability of each construction unit, and generate an optimized scheduling instruction set marking the risk hotspot; Feedback calibration module: used to obtain the actual execution data of the optimized scheduling instruction set, determine whether to trigger the third-level calibration instruction based on the actual execution data, and if triggered, feed back correction parameters to the perception terminal.

Images