A water conservancy project construction schedule dynamic correction system based on time series analysis

Abstract

The present application relates to the technical field of construction progress optimization, in particular to a water conservancy project construction progress dynamic correction system based on time sequence analysis, point cloud data of a construction excavation surface is obtained through a three-dimensional point cloud scanning device, structural feature information of the excavation surface is recognized and extracted, a structural mechanics model is generated, based on the structural mechanics model, the material strength evolution data in the reinforcing material are combined to perform multi-objective optimization on the support scheme and the operation time sequence, and a dynamic operation instruction is output. Based on the composite structure model containing the real rock mass structure and the real strength reinforcement system, the construction operation in the dynamic operation instruction is pre-calculated, the minimum safety margin of the composite structure is obtained, when the minimum safety margin is higher than the preset threshold, a construction authorization signal is output, and when the minimum safety margin is lower than the preset threshold, a construction prohibition signal is output and supplementary reinforcement is triggered. Therefore, the active intervention ability in the construction risk is improved.

Description

Technical Field

[0001] This invention relates to the field of construction progress optimization technology, specifically a dynamic correction system for the construction progress of water conservancy projects based on time series analysis. Background Technology

[0002] For construction schedule management of large-scale engineering projects, project management systems are typically used. Based on the work breakdown structure, these systems utilize classic project management theories such as the critical path method, program review and approval technique (PRP), or Gantt charts to plan, schedule, and allocate resources for each stage of construction. A global construction schedule is generated based on pre-defined task dependencies, durations, and required resources. Some systems also integrate resource management, cost control, and risk registration modules, allowing for manual updates to task status, recording of actual expenses, and tracking of identified risks during project execution.

[0003] However, these conventional project management systems have serious limitations when dealing with highly uncertain and dynamic engineering projects such as slope reinforcement of mountain reservoirs and dams. The progress of slope reinforcement construction in water conservancy projects is not only constrained by conventional personnel, equipment, and material supplies, but also heavily reliant on complex and dynamically changing on-site natural conditions, such as real-time monitoring of slope displacement data, changes in groundwater levels, and sudden severe weather. While existing systems can adjust plans, this adjustment typically requires managers to reassess the impact based on personal experience after problems are discovered, manually modifying system parameters and recalculating—a process that is inefficient and cannot meet the needs of real-time decision-making. There is a lack of a dynamic correction mechanism based on time-series data analysis, which cannot automatically and continuously integrate dynamic information flows from on-site sensor monitoring data and weather forecasts, nor can it predict potential schedule deviations and proactively generate corrective solutions based on this real-time data. Therefore, existing technologies cannot effectively solve the management problem of frequent construction plan failures caused by dynamic changes in the external environment.

[0004] To address this, a dynamic correction system for the construction progress of water conservancy projects based on time series analysis is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a dynamic correction system for the construction progress of water conservancy projects based on time-series analysis. This system acquires point cloud data of the excavation face using a 3D point cloud scanning device, identifies and extracts structural feature information of the excavation face, and generates a structural mechanics model. Based on this structural mechanics model, and combined with material strength evolution data in the reinforcement materials, the system performs multi-objective optimization of the support scheme and operation sequence, outputting dynamic operation instructions. Based on a composite structural model containing the actual rock mass structure and the actual strength reinforcement system, the system performs pre-calculation of the construction operations in the dynamic operation instructions, obtaining the minimum safety margin of the composite structure. When the minimum safety margin is higher than a preset threshold, a construction authorization signal is output; when the minimum safety margin is lower than the preset threshold, a construction prohibition signal is output, triggering supplementary reinforcement. This improves the proactive intervention capability in construction risks.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A dynamic correction system for the construction progress of water conservancy projects based on time series analysis includes:

[0008] The synchronous modeling module is used to acquire point cloud data of the construction excavation face through a 3D point cloud scanning device, and to use a pre-trained convolutional neural network model to identify and extract the structural feature information of the excavation face to generate a structural mechanics model.

[0009] The reinforcement optimization module is used to perform multi-objective optimization of the support scheme and operation sequence based on the structural mechanics model and the material strength evolution data in the reinforcement material, and output dynamic operation instructions.

[0010] The safety decision module is used to perform pre-calculation of construction operations in dynamic operation instructions based on a composite structure model that includes a real rock mass structure and a real strength reinforcement system, to obtain the minimum safety margin of the composite structure. When the minimum safety margin is higher than a preset threshold, a construction authorization signal is output; when the minimum safety margin is lower than the preset threshold, a construction prohibition signal is output and supplementary reinforcement is triggered.

[0011] Preferably, the structural mechanics model includes a geometric layer, a physical property layer, and a boundary condition layer;

[0012] The geometric layer adopts an adaptive mesh simplification method to convert the structural surface attitude angle, trace length, and aperture width data in the 3D point cloud data into simplified 3D geometric unit node coordinates and unit connection relationships, generating simplified discretized geometric units containing the main control joint surface and fracture surface;

[0013] The physical property layer uses a partitioned assignment method to assign the rock mass elastic modulus, Poisson's ratio, internal friction angle, cohesion data obtained from geological exploration, as well as the friction coefficient, normal stiffness, and tangential stiffness data of the structural surface, to the geometric unit according to the engineering partition, and sets the parameter uncertainty range to form a structural mechanics model.

[0014] The boundary condition layer uses a load grading algorithm to convert excavation unloading, ground stress, groundwater pressure, and construction load data into nodal force and displacement constraints of three-dimensional geometric unit nodes in stages. The geometric layer data provides a spatial positioning basis for the physical property layer, and the physical property layer data provides material response parameters for the boundary condition layer.

[0015] Preferably, the method for obtaining the material strength evolution data is as follows:

[0016] Sensor data from the anchor bolt grouting material is acquired, and the collected sensor data is input into the maturity theory using a modified hydration reaction kinetic model to obtain hydration degree data.

[0017] By establishing an empirical relationship model between strength and hydration degree, the hydration degree data is converted into time-varying trend data of compressive strength, tensile strength, and elastic modulus, generating a material strength evolution trend curve with time as the horizontal axis and strength parameters as the vertical axis.

[0018] Preferably, the dynamic job instruction acquisition method is as follows:

[0019] The structural mechanics model and material strength evolution data are input into a multi-objective genetic algorithm optimization engine. A safety factor is set as the primary constraint, while construction cost and construction period are secondary optimization objectives. Optimization variables include the discrete range of anchor bolt length, the discrete range of anchor bolt spacing, the limited type of grouting material, and the graded thickness of shotcrete. Strong constraints include a minimum safety factor greater than the specification requirement and material strength meeting design specification requirements. Using a weighted Pareto optimal solution screening method, the top three candidate solutions with the best comprehensive evaluation are selected from the set of constrained feasible solutions. Dynamic operation instructions are output, including construction procedure number, strength attainment trigger condition, construction parameter range, and completion judgment criteria.

[0020] Preferably, the composite structure model includes a rock mass sub-model, a support sub-model, and an interaction interface model, specifically:

[0021] The rock mass sub-model is based on data from the geometric units and physical property layers of the structural mechanics model, and uses elastoplastic constitutive relations and joint slip models to describe the mechanical behavior of the main rock mass and structural planes.

[0022] Based on the time-varying trend data in the material strength evolution data, the support sub-model uses simplified beam elements to simulate anchor bolts and simplified shell elements to simulate shotcrete, thereby obtaining the phased mechanical behavior of the support system.

[0023] The interaction interface model uses a linear contact algorithm to handle the contact force transmission between the rock mass and the support. Through deformation coordination conditions and static equilibrium conditions, a composite structure model of rock mass and support is formed. The rock mass sub-model data provides boundary conditions for the interaction interface, and the support sub-model data provides constraint parameters for the interaction interface.

[0024] Preferably, the minimum safety margin is obtained as follows:

[0025] Apply the load disturbance of the next excavation step to the composite structure model, perform structural response analysis, monitor the stress state of key units and nodes during the calculation process, and determine the dangerous state of rock mass units and support units.

[0026] The process for determining the hazardous state is as follows: calculate the safety factor of key elements and nodes, which is defined as the ratio of failure intensity to actual stress. Select the minimum safety factor of important control section elements as the minimum safety margin of the composite structure, and record the element location and failure mode corresponding to the minimum safety margin.

[0027] Preferably, the specific process of triggering supplementary reinforcement is as follows:

[0028] When the minimum safety margin is lower than a preset threshold, locate the cell position corresponding to the minimum safety margin and send an early warning message;

[0029] Based on the failure mode, corresponding reinforcement alternatives are recommended from the pre-set standardized reinforcement measures library, including the suggested location coordinates and technical parameter ranges for adding anchor bolts, the suggested area range and thickness parameter range for increasing shotcrete, and the suggested length and tension parameter range for setting prestressed anchor cables.

[0030] The selected reinforcement scheme data is fed back to the reinforcement optimization module for recalculation, dynamic operation instructions are updated, and the pre-calculation is repeated until the result meets the minimum safety margin requirement of being greater than the preset threshold.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] 1. This invention acquires point cloud data of the construction excavation face by introducing 3D point cloud scanning technology, intelligently extracts structural feature information, and assembles it into a digital structural mechanics model. This effectively improves the realism of the structural model at the construction site. The structural mechanics model includes a geometric layer, a physical property layer, and a boundary condition layer. The geometric layer uses an adaptive mesh simplification method to simplify the structural surfaces of the point cloud. The physical property layer assigns regional geological parameters to nodes and introduces a parameter uncertainty management mechanism. The boundary condition layer dynamically applies constraint mechanical conditions through a load grading algorithm, ensuring the model's responsiveness and accuracy to changes in working conditions at different stages. This accurately simulates the mechanical evolution process at the construction site, effectively improving the adaptability and feasibility of the project.

[0033] 2. This invention constructs a multi-objective optimization mechanism based on material strength evolution data through a reinforcement optimization module. It couples the formulation of support schemes with the time-varying strength behavior of construction materials, creating a dynamic operation instruction of "strength triggering condition + construction procedure number". Unlike rigid scheduling mechanisms based on empirical parameters, this invention can dynamically adjust the operation content according to the real-time construction progress and material properties, improving the efficiency of construction resource allocation and the flexibility of construction response.

[0034] 3. This invention constructs a closed-loop mechanism for construction safety evaluation and response through a dynamic safety assurance system consisting of a "composite structure model + minimum safety margin analysis + reinforcement strategy recommendation." Structural mechanics simulations are performed on the rock mass, support, and interaction interfaces. Using deformation coordination and contact force transfer modeling methods, the stress response of the composite structure is dynamically calculated after the next excavation load disturbance, and the minimum safety factor of key units is extracted as a safety margin index. This enables high-frequency, real-time, model-driven safety prediction and reinforcement scheduling, improving the system's proactive intervention capability and decision-making intelligence level in the face of sudden construction risks, thereby enhancing the stability of the entire water conservancy project construction phase. Attached Figure Description

[0035] Figure 1 A schematic diagram of a dynamic correction system for construction progress of water conservancy projects based on time series analysis is provided for this invention.

[0036] Figure 2 A schematic diagram of the dynamic correction process for the construction progress of water conservancy projects based on time-series analysis provided by the present invention;

[0037] Figure 3 This is a schematic diagram of the dynamic correction process for construction progress provided in an embodiment of the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Example 1:

[0040] This invention provides a dynamic correction system for the construction progress of water conservancy projects based on time series analysis, with a framework structure referencing... Figure 1 A schematic diagram of a dynamic correction system for construction progress in water conservancy projects based on time-series analysis is provided. The specific implementation process is as follows: Figure 2 The details are as follows:

[0041] The synchronous modeling module is used to acquire point cloud data of the construction excavation face through a 3D point cloud scanning device, and to identify and extract the structural feature information of the excavation face using a pre-trained convolutional neural network model. The structural feature information is then assembled into a digital mechanics model to generate a structural mechanics model.

[0042] The structural mechanics model includes a geometric layer, a physical property layer, and a boundary condition layer;

[0043] The geometric layer adopts an adaptive mesh simplification method to convert the structural surface attitude angle, trace length, and aperture width data in the 3D point cloud data into simplified 3D geometric unit node coordinates and unit connection relationships, generating simplified discretized geometric units containing the main control joint surface and fracture surface;

[0044] The physical property layer uses a partitioned assignment method to assign the rock mass elastic modulus, Poisson's ratio, internal friction angle, cohesion data obtained from geological exploration, as well as the friction coefficient, normal stiffness, and tangential stiffness data of the structural surface, to the geometric unit according to the engineering partition, and sets the parameter uncertainty range to form a structural mechanics model.

[0045] The boundary condition layer uses a load grading algorithm to convert excavation unloading, ground stress, groundwater pressure, and construction load data into nodal force and displacement constraints of three-dimensional geometric unit nodes in stages. The geometric layer data provides a spatial positioning basis for the physical property layer, and the physical property layer data provides material response parameters for the boundary condition layer.

[0046] In this embodiment, a synchronous modeling module, utilizing a 3D point cloud scanning device and a convolutional neural network model, enables real-time and accurate acquisition of structural feature information of the construction excavation face. This information is then efficiently integrated into a digital mechanics model, generating a isomorphic structural mechanics model, significantly improving model accuracy and reliability. The geometric layer employs an adaptive mesh simplification method, effectively reducing computational load and improving efficiency. The physical attribute layer uses a partitioned assignment method to fully consider the uncertainties of geological conditions, enhancing model accuracy. The boundary condition layer utilizes a load-grading algorithm to realistically simulate changes in construction loads, improving model applicability. This solution provides a scientific basis for engineering design and construction, offering advantages of high efficiency, accuracy, and reliability.

[0047] The reinforcement optimization module is used to perform multi-objective optimization of the support scheme and operation sequence based on the structural mechanics model and the material strength evolution data in the reinforcement material, and output dynamic operation instructions.

[0048] The method for obtaining the material strength evolution data is as follows:

[0049] Sensor data from the anchor bolt grouting material is acquired, and the collected sensor data is input into the maturity theory using a modified hydration reaction kinetic model to obtain hydration degree data. The sensors include a wireless temperature sensor, a humidity sensor, and a resistivity sensor. The sensor sampling frequency is once every 3 hours, and a periodic calibration mechanism is set.

[0050] The modified hydration reaction kinetic model includes an environmental correction layer, a reaction rate calculation layer, and a hydration degree prediction layer:

[0051] The environmental correction layer compares and analyzes the measured temperature, humidity, and environmental pressure data collected by the sensors with the baseline parameters under standard maintenance conditions. Using the temperature correction coefficient calculation formula and the humidity influence factor calculation formula, an environmental correction parameter set containing temperature correction coefficient, humidity correction coefficient, and pressure correction coefficient is generated. The environmental correction parameter set is used to correct the theoretical hydration reaction rate.

[0052] The reaction rate calculation layer uses environmental correction parameter set, material composition data, water-cement ratio data, and admixture type data as input variables to calculate the actual hydration reaction rate curve and cumulative hydration reaction degree curve considering the influence of the on-site environment.

[0053] The hydration degree prediction layer employs an integral calculation method, numerically integrating the actual hydration reaction rate curve over time steps. Combined with cement clinker mineral composition data and the formation patterns of hydration products, a hydration degree evolution curve with time as the independent variable is generated, thus obtaining hydration degree data. This hydration degree evolution curve data provides core input parameters for the strength-hydration degree empirical relationship model.

[0054] By establishing an empirical relationship model between strength and hydration degree, the hydration degree data is converted into time-varying trend data of compressive strength, tensile strength, and elastic modulus, generating a material strength evolution trend curve with time as the horizontal axis and strength parameters as the vertical axis.

[0055] The intensity-hydration degree empirical relationship model includes an intensity prediction layer, a parameter correction layer, and a reliability assessment layer:

[0056] The strength prediction layer inputs hydration degree evolution curve data, material grade data, and mix design data into the strength development database. Through the strength-hydration degree relationship function and multiple regression analysis method, it calculates the theoretical predicted values ​​of the time-varying curves of compressive strength, tensile strength, and elastic modulus.

[0057] The parameter correction layer is based on field sampling and testing data. It performs error analysis between the measured material strength data and the theoretical prediction value, calculates the compressive strength correction factor, tensile strength correction factor, and elastic modulus correction factor, and applies the correction factors to the theoretical prediction value to generate the corrected material strength evolution trend curve.

[0058] The reliability assessment layer uses statistical analysis methods to compare and analyze historical test data, data from similar projects, and standard test data to calculate upper limit, lower limit, and confidence index for each strength parameter. The material strength evolution trend curve and reliability index data provide the reinforcement optimization module with material parameter input and uncertainty quantification basis.

[0059] In this embodiment, sensors collect real-time data on the temperature, humidity, and resistivity of the anchor grouting material. A modified hydration reaction kinetic model is used to generate a hydration degree evolution curve, which is then converted into a time-varying trend of material strength, providing accurate input for reinforcement optimization. An environmental correction layer considers the influence of on-site temperature, humidity, and pressure, improving model adaptability and prediction accuracy. The reinforcement optimization module combines strength data for multi-objective optimization, outputting dynamic operation instructions to optimize the support scheme. The strength-hydration degree model ensures reliable predictions through parameter correction and reliability assessment. High-frequency sensor sampling and periodic calibration support intelligent management. This improves the dynamic optimization level of the support and enhances engineering safety.

[0060] The method for obtaining the dynamic job instruction is as follows:

[0061] The structural mechanics model and material strength evolution data are input into a multi-objective genetic algorithm optimization engine. A safety factor is set as the primary constraint, while construction cost and construction period are secondary optimization objectives. Optimization variables include the discrete range of anchor bolt length, the discrete range of anchor bolt spacing, the limited type of grouting material, and the graded thickness of shotcrete. Strong constraints include a minimum safety factor greater than the specification requirement and material strength meeting design specification requirements. Using a weighted Pareto optimal solution screening method, the top three candidate solutions with the best comprehensive evaluation are selected from the set of constrained feasible solutions. Dynamic operation instructions are output, including construction procedure number, strength attainment trigger condition, construction parameter range, and completion judgment criteria.

[0062] The multi-objective genetic algorithm optimization engine includes a population initialization layer, a fitness evaluation layer, and an evolutionary operation layer:

[0063] The population initialization layer uses the discrete range data of anchor bolt length, discrete range data of anchor bolt spacing, data on the limited type of grouting material, and data on the graded thickness of shotcrete as the gene coding space to generate initial population individuals containing anchor bolt length gene, spacing gene, material type gene, and thickness gene;

[0064] The fitness evaluation layer takes structural mechanics model data, material strength evolution data, engineering cost database, and construction period quota data as inputs. Through safety factor calculation algorithm, cost estimation algorithm, and construction period prediction algorithm, it calculates safety factor evaluation value, construction cost evaluation value, and construction period evaluation value for each individual in the population. It also uses constraint violation degree calculation method to handle strong constraint conditions.

[0065] The evolutionary operation layer performs Pareto stratification on the fitness evaluation results using a non-dominated sorting algorithm and a crowding distance calculation algorithm. It then uses tournament selection, single-point crossover, and uniform mutation to generate a new generation of population. The output data of the evolutionary operation layer is a Pareto optimal solution set optimized through multiple generations of evolution.

[0066] In this embodiment, a multi-objective genetic algorithm is used to optimize construction parameters, with the safety factor as the primary constraint, while also minimizing cost and construction period. By integrating structural mechanics models and material strength evolution data, dynamic work instructions are generated, improving construction accuracy and efficiency. Advanced evolutionary techniques ensure diverse and high-quality solution sets, providing flexibility for decision-making.

[0067] The safety decision module is used to perform pre-calculation of construction operations in dynamic operation instructions based on a composite structure model that includes a real rock mass structure and a real strength reinforcement system, to obtain the minimum safety margin of the composite structure. When the minimum safety margin is higher than a preset threshold, a construction authorization signal is output; when the minimum safety margin is lower than the preset threshold, a construction prohibition signal is output and supplementary reinforcement is triggered.

[0068] The composite structure model includes a rock mass sub-model, a support sub-model, and an interaction interface model, specifically:

[0069] The rock mass sub-model is based on data from the geometric units and physical property layers of the structural mechanics model, and uses elastoplastic constitutive relations and joint slip models to handle the mechanical behavior of the main rock mass and structural planes;

[0070] Based on the time-varying trend data in the material strength evolution data, the support sub-model uses simplified beam elements to simulate anchor bolts and simplified shell elements to simulate shotcrete, thereby obtaining the phased mechanical behavior of the support system.

[0071] The interaction interface model uses a linear contact algorithm to handle the contact force transmission between the rock mass and the support. Through deformation coordination conditions and static equilibrium conditions, a composite structure model of rock mass and support is formed. The rock mass sub-model data provides boundary conditions for the interaction interface, and the support sub-model data provides constraint parameters for the interaction interface.

[0072] The joint surface slip model includes a geometric description layer, a constitutive relation layer, and a slip determination layer:

[0073] The geometric description layer converts the joint surface attitude angle data, trace length dimension data, and aperture width data extracted from the 3D point cloud data into the normal vector coordinates, center point coordinates, and boundary contour coordinates of the joint surface through coordinate transformation algorithms and geometric projection algorithms, and establishes a spatial geometric description model of the joint surface.

[0074] The constitutive relation layer takes the joint surface friction coefficient data, normal stiffness data, tangential stiffness data, and roughness coefficient data as material parameters as inputs, and establishes a nonlinear constitutive relation model of the joint surface considering normal deformation and tangential slip through the calculation of normal stress-displacement relationship and tangential stress-displacement relationship.

[0075] The slip determination layer compares the normal stress data and tangential stress data on the joint surface with the shear strength parameters of the joint surface. Through the slip condition determination algorithm and the slip direction calculation algorithm, it determines the slip state and slip displacement increment of the joint surface. The output data of the slip determination layer provides the joint surface mechanical response parameters and deformation boundary conditions for the rock mass sub-model.

[0076] In this embodiment, a composite structural model incorporating a realistic rock mass structure and a reinforcement system effectively enhances construction safety. The composite structural model integrates a rock mass sub-model, a support sub-model, and an interaction interface model to simulate the mechanical behavior of the rock mass and support, ensuring stability. The joint surface slip model refines stress and slip analysis and optimizes safety assessment through geometric description, constitutive relations, and slip determination layers.

[0077] The minimum safety margin is obtained as follows:

[0078] Apply the load disturbance of the next excavation step to the composite structure model, perform structural response analysis, monitor the stress state of key units and nodes during the calculation process, and determine the dangerous state of rock mass units and support units.

[0079] The process for determining the hazardous state is as follows: calculate the safety factor of key elements and nodes, which is defined as the ratio of failure intensity to actual stress. Select the minimum safety factor of important control section elements as the minimum safety margin of the composite structure, and record the element location and failure mode corresponding to the minimum safety margin.

[0080] In this embodiment, by applying load disturbances to the composite structure model and performing stress analysis, potential failure points are predicted, thereby improving construction safety. Determining the minimum safety margin allows for the pre-identification and reinforcement of weak points, preventing accidents and reducing subsequent repair costs. Simultaneously, recording failure locations and modes optimizes resource allocation and enhances the targetedness and efficiency of support design.

[0081] The specific process for triggering supplementary reinforcement is as follows:

[0082] When the minimum safety margin is lower than a preset threshold, locate the cell position corresponding to the minimum safety margin and send an early warning message;

[0083] Based on the failure mode, corresponding reinforcement alternatives are recommended from the pre-set standardized reinforcement measures library, including the suggested location coordinates and technical parameter ranges for adding anchor bolts, the suggested area range and thickness parameter range for increasing shotcrete, and the suggested length and tension parameter range for setting prestressed anchor cables.

[0084] The selected reinforcement scheme data is fed back to the reinforcement optimization module for recalculation, dynamic operation instructions are updated, and the pre-calculation is repeated until the result meets the minimum safety margin requirement of being greater than the preset threshold.

[0085] In this embodiment, weak elements are located and early warnings are issued to prevent potential risks in a timely manner. Targeted solutions are recommended from a standardized reinforcement measures library, such as adding anchor bolts, increasing the thickness of shotcrete, or installing prestressed anchor cables, with precise parameters provided to ensure efficient and economical reinforcement. The selected solution is fed back to the optimization module for iterative calculations until the safety margin is met, ensuring structural safety while optimizing resource allocation and reducing costs.

[0086] This invention provides a system integrating real-time isomorphic modeling, reinforcement optimization, and safety decision-making. The synchronous modeling module acquires the structural features of the excavation face through 3D point cloud scanning and convolutional neural networks, generating an isomorphic structural mechanical model. Adaptive mesh simplification reduces computational load, while geological uncertainty assignment and load grading enhance model reliability, providing a scientific basis for design and construction. The reinforcement optimization module utilizes sensor data and a modified hydration reaction kinetic model to predict material strength evolution. It combines a multi-objective genetic algorithm to optimize support schemes and operational sequences, outputting dynamic commands to improve safety and efficiency. Environmental correction and parameter calibration ensure accurate predictions. The safety decision-making module pre-simulates construction operations based on the composite structure model, calculates the minimum safety margin, triggers targeted reinforcement, and reduces accident risks and repair costs. (See details...) Figure 3 The joint surface slip model refines mechanical analysis and optimizes resource allocation. This system integrates advanced technologies to achieve comprehensive optimization of safety, cost, and schedule, thus promoting advancements in construction management technology.

[0087] Example 2:

[0088] As another specific embodiment of the present invention, taking the left bank slope reinforcement project of a reservoir dam as an application scenario, the specific implementation process of the present invention is described in detail. The rock mass is mainly granite, with three main joint surfaces, requiring combined support of anchor bolts and shotcrete.

[0089] In one embodiment of the present invention, the synchronous modeling module uses a 3D laser scanner to scan the excavated slope surface. The point cloud density is set to one point per square centimeter to ensure that the acquired point cloud data accurately reflects the detailed features of the slope surface. A scanning station is set up 30 meters in front of the slope, using a multi-station combined scanning method. Each scanning station covers a 120-degree area, and the overlap between adjacent stations is no less than 30%, ensuring the integrity and continuity of the point cloud data.

[0090] In one embodiment of the present invention, the pre-trained convolutional neural network model employs an image recognition algorithm based on a deep residual network architecture. This network contains 50 convolutional layers and uses weight parameters pre-trained on a rock engineering image database. The training database contains rock surface images under different geological conditions, each image annotated with structural features such as the attitude angle, trace length, and opening width of the joint surfaces. The network model can identify the dip angle range, dip angle magnitude, extension length, and opening degree of the joint surfaces. In this embodiment, three main groups of joint surfaces are identified: the first group dips at 45 degrees northeast with a dip angle of 70 degrees; the second group dips at 60 degrees northwest with a dip angle of 65 degrees; and the third group dips due north with a dip angle of 85 degrees.

[0091] The adaptive mesh simplification method for the geometric layers dynamically adjusts the mesh density based on the complexity of the rock surface. In areas with dense joints, the mesh size is set to 0.2 meters, while in relatively intact rock areas, the mesh size is increased to 1.5 meters. The mesh simplification criterion is based on the rate of curvature change of adjacent point clouds; mesh refinement is performed when the rate of curvature change is greater than 0.15, and mesh coarsening is performed when the rate of curvature change is less than 0.05. The final generated 3D geometric unit contains 8500 nodes and 15000 tetrahedral elements, with skewness all less than 0.6 in the mesh quality evaluation index.

[0092] The physical property layer assigns values ​​to the rock mass parameters obtained from geological exploration according to engineering zones. The parameters of the first, second, and third groups of joint surfaces are determined based on the results of field sampling tests. Considering the uncertainty of geological parameters, the system sets upper and lower limits for each parameter: the elastic modulus varies within ±20% of the standard value, and the strength parameter varies within ±15% of the standard value.

[0093] The boundary condition layer simulates the step-by-step excavation and support process using a load-grading algorithm. Excavation and unloading follow a top-down, layered excavation sequence. Ground stress is applied based on the measured original rock stress state, with a horizontal principal stress of 1.2 MPa and a vertical principal stress of 2.1 MPa. Groundwater pressure is applied based on water level monitoring data, following the hydrostatic pressure distribution pattern. Construction loads include the dynamic loads of the excavation equipment and the static loads of material stacking, which are converted into load boundary conditions for the corresponding nodes.

[0094] In one embodiment of the present invention, material strength evolution data is acquired in real time by sensors embedded in the anchor bolt grouting body. The sensor system includes a wireless temperature sensor, a humidity sensor, and a resistivity sensor, which are installed in the anchor bolt grouting holes at different locations. The temperature sensor has a measurement accuracy of ±0.2 degrees Celsius, the humidity sensor has a measurement accuracy of ±3% relative humidity, and the resistivity sensor has a measurement accuracy of ±2%. The sensor sampling frequency is set to once every 3 hours.

[0095] In one embodiment of the present invention, the modified hydration reaction kinetic model considers the influence of on-site environmental conditions on the hydration reaction rate. The environmental correction layer compares the measured temperature data with the standard curing temperature of 20 degrees Celsius. When the measured temperature is 15 degrees Celsius, the temperature correction factor is 0.7; when the measured temperature is 25 degrees Celsius, the temperature correction factor is 1.3. The humidity correction layer compares the measured humidity with the standard humidity of 95%. The humidity correction factor decreases as humidity decreases, reaching 0.9 when the humidity is 80%. The pressure correction layer considers the influence of altitude on atmospheric pressure. These correction factors were obtained through historical data and expert experience.

[0096] In one embodiment of the present invention, the hydration degree prediction layer calculates the degree of hydration reaction using a numerical integration method. The reaction rate calculation layer comprehensively considers factors such as environmental correction parameters, cement composition data, water-cement ratio, and water-reducing agent dosage to generate an actual hydration reaction rate curve that takes into account on-site conditions. The integration calculation uses the trapezoidal integration method with a time step of 1 hour, and combines the mineral composition data of silicate cement and the formation rules of hydration products to generate a hydration degree evolution curve with time as the variable.

[0097] In one embodiment of the present invention, the strength-hydration degree empirical relationship model converts hydration degree data into material strength parameters. The strength prediction layer calculates the time-varying curves of compressive strength, tensile strength, and elastic modulus based on the hydration degree evolution curve, cement grade data, and mix design data, through strength development database queries and multiple regression analysis. The parameter correction layer performs error analysis using strength test data from cubic specimens sampled on-site. The reliability assessment layer determines the confidence interval for each strength parameter by comparing historical engineering data and standard test data, with a 95% confidence level and a 12% coefficient of variation for compressive strength.

[0098] In one embodiment of the present invention, dynamic operation instructions are obtained through optimization using a multi-objective genetic algorithm. The optimization engine's population initialization layer sets the anchor bolt length to a discrete range of 6 to 12 meters, the spacing to a discrete range of 1.5 to 3.0 meters, and the grouting material type to include either ordinary silicate cement grout or cement-water glass dual-liquid grout. The initial population size is set to 100 individuals, each individual containing four genes: anchor bolt length, spacing, material type, and thickness.

[0099] The fitness evaluation layer comprehensively considers three objectives: safety, economy, and timeliness. The safety factor is calculated based on the results of structural mechanics analysis; cost estimation includes material costs, labor costs, and equipment costs; the construction period prediction is based on construction quotas; the evolutionary operation layer adopts tournament selection, single-point crossover, and uniform mutation operations, and converges to the Pareto optimal solution set after 200 generations of evolution.

[0100] The final output of the dynamic operation instruction scheme includes detailed construction process numbers, strength attainment trigger conditions, construction parameter ranges, and completion judgment criteria.

[0101] In one embodiment of the present invention, the composite structural model integrates three sub-models: rock mass, support, and interaction interface. The rock mass sub-model, based on the geometric elements and physical property data of the structural mechanics model, uses an elastoplastic constitutive relation to describe the stress-strain behavior of the rock mass. When the rock mass stress exceeds the yield strength, the material enters a plastic state, and the strength parameters are reduced according to the Mohr-Coulomb failure criterion. The joint surface slip model establishes the spatial relationship of joint surfaces through a geometric description layer, determines the mechanical response based on the friction coefficient, normal stiffness, and tangential stiffness of the joint surfaces through a constitutive relation layer, and determines whether slip has occurred based on a comparison of shear stress and shear strength through a slip determination layer.

[0102] The support sub-model uses beam elements to simulate the axial force and bending deformation of the anchor rods, with the rod diameter set at 25 mm and the material strength determined according to the steel reinforcement parameters; shell elements simulate the internal force and bending deformation of the shotcrete; the time-varying characteristics of the support system are dynamically updated through material strength evolution data, and the strength parameters of the anchor grout body are adjusted in real time at different ages.

[0103] The interaction interface model handles the transmission of contact forces between the rock mass and the support. When the rock mass and the support undergo relative displacement, the contact force is calculated based on the contact stiffness. Deformation compatibility conditions ensure the continuity of displacement on the contact surface, and static equilibrium conditions ensure that the transmission of force satisfies Newton's third law.

[0104] In one embodiment of the present invention, the minimum safety margin is obtained by applying the next excavation load disturbance to the composite structure model. The structural response analysis employs the nonlinear finite element method, considering the effects of material nonlinearity and geometric nonlinearity. During the calculation process, the stress state of key elements within a 5-meter radius of the excavation face, including rock mass elements and support elements, is closely monitored.

[0105] The hazard determination process consists of three levels. First, the principal stress state of each element is calculated to determine whether the tensile stress exceeds the tensile strength and whether the compressive stress exceeds the compressive strength. Then, the shear stress state is calculated to determine whether the Mohr-Coulomb failure criterion is satisfied. Finally, considering the three failure modes of tension, compression, and shear, the corresponding safety factors are calculated. The tensile safety factor is defined as the ratio of tensile strength to maximum tensile stress; the compressive safety factor is defined as the ratio of compressive strength to maximum compressive stress; and the shear safety factor is defined as the ratio of shear strength to maximum shear stress.

[0106] The system selects the minimum safety factor among all critical elements as the minimum safety margin of the composite structure, and records the element number, spatial coordinates and failure mode type corresponding to the minimum value.

[0107] In one embodiment of the present invention, a supplementary reinforcement procedure is triggered when the minimum safety margin falls below a preset threshold. Warning information is simultaneously sent to project management personnel via a mobile terminal and an on-site display screen. The warning content includes the specific location of the hazardous unit, the safety margin value, the estimated time of failure, and recommended emergency measures.

[0108] After selecting a reinforcement scheme, the system feeds the scheme data back to the reinforcement optimization module for recalculation based on multi-objective objectives. The optimization process considers the material strength evolution characteristics of the newly added reinforcement measures and updates the construction parameters and timing in the dynamic work instructions. The safety decision module repeatedly performs pre-calculations on the updated scheme to verify whether the reinforcement effect meets safety requirements. After two rounds of iterative calculations, the minimum safety margin improvement after adopting the combined reinforcement scheme meets the safety threshold requirement of greater than 1.5, and a construction authorization signal is output, allowing the next excavation operation to proceed.

[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A water conservancy construction progress dynamic correction system based on timing analysis, characterized in that, The method comprises the following steps: A synchronous modeling module is used to obtain point cloud data of a construction excavation surface through a three-dimensional point cloud scanning device, and a pre-trained convolutional neural network model is used to identify and extract structural feature information of the excavation surface to generate a structural mechanics model; A reinforcement optimization module is used to perform multi-objective optimization on a support scheme and an operation time sequence based on the structural mechanics model and in combination with material strength evolution data in a reinforcing material, and to output dynamic operation instructions; The dynamic operation instruction is obtained in the following manner: The structural mechanics model and the material strength evolution data are input into a multi-objective genetic algorithm optimization engine, a safety factor is set as a primary constraint condition, and construction cost and construction duration are set as secondary optimization objectives; optimization variables include an anchor rod length discrete range, an anchor rod distance discrete range, a grouting material limited type, and a shotcrete thickness classification, and strong constraint conditions include a minimum safety factor greater than a specification requirement value and material strength meeting a design specification requirement; the first three selected schemes with the best comprehensive evaluation are selected from a constraint feasible solution set through a weighted Pareto optimal solution screening method, and dynamic operation instructions containing a construction procedure number, a strength compliance trigger condition, a construction parameter range, and a completion determination standard are output; A safety decision module is used to perform pre-play calculation on construction operations in the dynamic operation instructions based on a composite structure model containing a real rock mass structure and a real strength reinforcement system, to obtain a minimum safety margin of the composite structure, to output a construction authorization signal when the minimum safety margin is higher than a preset threshold, and to output a construction prohibition signal and trigger supplementary reinforcement when the minimum safety margin is lower than the preset threshold.

2. The water conservancy construction schedule dynamic correction system based on timing analysis according to claim 1, characterized in that: The structural mechanics model comprises a geometric layer, a physical attribute layer, and a boundary condition layer; The geometric layer uses an adaptive mesh simplification method to convert structure surface occurrence angle, trace length size, and opening width data in the three-dimensional point cloud data into simplified three-dimensional geometric unit node coordinates and unit connection relationships, to generate simplified discrete geometric units containing a master joint surface and a fissure surface; The physical attribute layer uses a partition assignment method to assign rock mass elastic modulus, Poisson's ratio, internal friction angle, and cohesion data obtained through geological exploration and structure surface friction coefficient, normal stiffness, and tangential stiffness data to the geometric units according to engineering partitions, and sets a parameter uncertainty range to form the structural mechanics model; The boundary condition layer uses a load grading application algorithm to convert excavation unloading, ground stress, underground water pressure, and construction load data into node force and displacement constraint conditions of the three-dimensional geometric unit nodes in stages, and the geometric layer data provides a spatial positioning basis for the physical attribute layer, and the physical attribute layer data provides material response parameters for the boundary condition layer.

3. The water conservancy construction schedule dynamic correction system based on time series analysis according to claim 1, characterized in that: The material strength evolution data is obtained in the following manner: Sensor data in an anchor rod grouting material is obtained, and a modified hydration reaction kinetics model is used to input the collected sensor data into a maturity theory to obtain hydration degree data; An established strength-hydration degree empirical relationship model is used to convert the hydration degree data into time-varying trend data of compressive strength, tensile strength, and elastic modulus to generate a material strength evolution trend curve with time as the horizontal axis and strength parameters as the vertical axis.

4. The water conservancy construction schedule dynamic correction system based on timing analysis according to claim 1, characterized in that: The composite structure model comprises a rock mass submodel, a support submodel and an interaction interface model, and specifically comprises: The rock mass submodel is based on the data in the geometric element and physical attribute layer of the structural mechanics model, adopts an elastic-plastic constitutive relation and a joint surface slip model to process the mechanical behavior of the main rock mass and structural surface; The support submodel adopts a simplified beam element to simulate the anchor rod and a simplified shell element to simulate the shotcrete according to the time-varying trend data in the material strength evolution data, and obtains the staged mechanical behavior of the support system; The interaction interface model uses a linear contact algorithm to process the contact force transmission between the rock mass and the support, forms a rock mass-support composite structure model through deformation coordination conditions and static force balance conditions, and provides boundary conditions for the interaction interface with the data of the rock mass submodel and provides constraint parameters for the interaction interface with the data of the support submodel.

5. The water conservancy construction schedule dynamic correction system based on time series analysis of claim 1, wherein: The minimum safety margin is obtained as follows: The load disturbance of the next step of excavation is applied to the composite structure model to perform structural response analysis, and the stress state of the key elements and nodes is monitored in the calculation process to determine the dangerous state of the rock mass element and the dangerous state of the support element; The dangerous state determination process is as follows: the safety factor of the key elements and nodes is calculated, which is defined as the ratio of the failure strength to the actual stress, the minimum value of the safety factor of the important control section element is selected as the minimum safety margin of the composite structure, and the element position and failure mode corresponding to the minimum safety margin are recorded.

6. The water conservancy construction schedule dynamic correction system based on timing analysis according to claim 1, characterized in that: The specific process of triggering supplementary reinforcement is as follows: When the minimum safety margin is lower than the preset threshold value, the element position corresponding to the minimum safety margin is located, and a warning information is sent; According to the failure mode, a corresponding reinforcement alternative scheme is recommended from the preset standardized reinforcement measure library, including the recommended position coordinates and technical parameter range of the additional anchor rod, the recommended area range and thickness parameter range of the additional shotcrete, and the recommended length and tension force parameter range of the pre-stressed anchor cable; The selected reinforcement scheme data is fed back to the reinforcement optimization module for re-optimization calculation, the dynamic operation instruction is updated, and the pre-rehearsal calculation is repeated until the result meets the condition that the minimum safety margin is greater than the preset threshold value.

Images