A Method and System for Locating Anchor Points in Cofferdam Bridge Construction Based on Digital Twin

Abstract

This invention provides a method and system for anchor point positioning in cofferdam bridge construction based on digital twins, belonging to the field of digital twin technology. The method includes: using a drone to acquire images and perform laser scanning of the cofferdam bridge construction area, fusing the image set and point cloud set to construct a three-dimensional construction twin model; randomly distributing anchor points based on a preset number of anchor points and generating a distribution; using the three-dimensional construction twin model, simulating external loads and construction based on the anchor point distribution, and evaluating the distribution's fitness; optimizing the anchor point distribution based on the fitness until convergence, and outputting the optimal anchor point distribution for construction. This invention solves the technical problems in traditional cofferdam bridge construction where anchor point positioning relies on manual experience and static design drawings, leading to large layout deviations and difficulty in adapting to dynamic construction disturbances. It achieves the technical effect of optimizing anchor point positions through a digital twin model, improving construction accuracy, reducing construction errors, and thus improving construction efficiency and safety.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, specifically to a method and system for locating anchor points in cofferdam bridge construction based on digital twins. Background Technology

[0002] Cofferdam bridge construction is a crucial component of modern bridge development, and its precision directly impacts the bridge's safety and stability. In traditional cofferdam bridge construction, the layout of anchorage points typically relies on manual experience and site conditions, leading to uneven distribution, significant construction errors, and the inability to monitor and adjust in real-time, resulting in unstable construction quality. Furthermore, the construction environment is frequently influenced by various factors such as external loads, construction progress, and environmental changes, making precise control difficult with traditional methods. With the development of digital technology, digital twin technology, as an innovative engineering management tool, is increasingly being applied to bridge construction. However, existing technologies still struggle to respond promptly to changes in the construction environment and to correct errors during dynamic construction simulation and optimization. Therefore, a highly efficient anchorage point positioning method based on digital twins is urgently needed to overcome the limitations of traditional construction positioning techniques in terms of accuracy maintenance, environmental adaptability, and safety redundancy. Summary of the Invention

[0003] This application provides a method and system for anchor point positioning in cofferdam bridge construction based on digital twins, aiming to solve the technical problems of large layout deviations and difficulty in adapting to dynamic construction disturbances caused by the reliance on manual experience and static design drawings for anchor point positioning in traditional cofferdam bridge construction.

[0004] The first aspect disclosed in this application provides a method for locating anchorage points in cofferdam bridge construction based on digital twins. The method includes: using a drone to acquire multi-angle images and perform laser scanning of the cofferdam bridge construction area; constructing a three-dimensional construction twin model by fusing regional image sets and regional point clouds; randomly arranging anchorage points within the cofferdam bridge construction area based on a preset number of anchorage points, generating several anchorage point distributions; using the three-dimensional construction twin model, performing external load simulation and construction simulation based on the several anchorage point distributions, and evaluating and determining the fitness of several distributions based on the simulation results; iteratively optimizing the anchorage point distribution based on the several distribution fitnesss until convergence, outputting the optimal anchorage point distribution for anchorage point construction.

[0005] The second aspect of this application discloses a digital twin-based anchorage point positioning system for cofferdam bridge construction. This system is used in the aforementioned digital twin-based anchorage point positioning method for cofferdam bridge construction. The system includes: a twin modeling module: using a drone to acquire multi-angle images and perform laser scanning of the cofferdam bridge construction area, and constructing a three-dimensional construction twin model based on the fusion of the regional image set and regional point cloud set; an anchorage point layout module: randomly arranging anchorage points within the cofferdam bridge construction area based on a preset number of anchorage points, generating several anchorage point distributions; a model simulation module: using the three-dimensional construction twin model, performing external load simulation and construction simulation based on the several anchorage point distributions, and evaluating and determining the fitness of several distributions based on the simulation results; and an anchorage point construction module: iteratively optimizing the anchorage point distribution based on the several distribution fitnesss until convergence, outputting the optimal anchorage point distribution for anchorage point construction.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] This method utilizes drones to capture multi-angle images and perform laser scanning of the cofferdam bridge construction area. Combining the obtained image data and point cloud data, a three-dimensional digital twin model is constructed. Subsequently, within the construction area, anchor points are randomly arranged according to a predetermined number, generating different distributions. Then, using the constructed 3D model, external load and construction simulations are performed on each anchor point distribution to evaluate its suitability. Finally, by comparing the suitability of different distributions, iterative optimization is performed until the optimal anchor point distribution is found. Anchor point construction is then carried out according to this distribution, thereby improving construction efficiency and safety.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 A schematic diagram of the process for locating anchor points in cofferdam bridge construction based on digital twins, provided in an embodiment of this application.

[0010] Figure 2 A schematic diagram of the structure of the cofferdam bridge construction anchorage point positioning system based on digital twin provided in this application embodiment.

[0011] Figure labeling: Twin modeling module 11, anchor point layout module 12, model simulation module 13, anchor point construction module 14. Detailed Implementation

[0012] This application provides a method and system for locating anchor points in cofferdam bridge construction based on digital twins, which solves the technical problems of large layout deviations and difficulty in adapting to dynamic construction disturbances caused by the reliance on manual experience and static design drawings for anchor point positioning in traditional cofferdam bridge construction.

[0013] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0014] Example 1, as Figure 1 As shown in the embodiment of this application, a method for locating anchorage points in cofferdam bridge construction based on digital twins is provided. The method includes:

[0015] Using drones, multi-angle image acquisition and laser scanning were conducted on the construction area of ​​the cofferdam bridge. A three-dimensional construction twin model was constructed by fusing the regional image set and the regional point cloud set.

[0016] In this embodiment, a high-definition camera and laser scanner mounted on a drone are used to capture images and perform laser scanning from multiple angles in the cofferdam bridge construction area. This comprehensively acquires high-resolution images and precise 3D point cloud data of the construction area, forming a regional image set and a regional point cloud set. Subsequently, an edge detection algorithm is used to extract the contours from the regional image set, and then the contour extraction results are used for 3D reconstruction to determine the regional 3D contour of the cofferdam bridge construction area. Then, by fusing the regional 3D contour with the regional point cloud set, a detailed 3D construction twin model is constructed. This 3D construction twin model realistically reflects the geometry and physical characteristics of the construction site, providing accurate virtual environment support for subsequent construction planning, monitoring, and optimization, thereby avoiding errors that may occur in traditional manual measurements.

[0017] Furthermore, a 3D construction twin model is constructed by fusing regional image sets and regional point cloud sets, including:

[0018] The region image set is extracted using an edge detection algorithm to obtain a region contour set; a three-dimensional reconstruction is performed based on the region contour set according to a preset acquisition angle to output the region three-dimensional contour; the region point cloud set is fitted with the region point cloud set based on the region three-dimensional contour to obtain the optimal point cloud fitting result and construct a three-dimensional construction twin model.

[0019] Preferably, after obtaining the region image set and region point cloud, the region image set is first processed using an edge detection algorithm (such as Canny edge detection or the Sobel operator) to identify edge components in the image. Taking Canny edge detection as an example, a Gaussian filter is first applied to smooth the image and reduce noise interference. Then, potential edge regions are identified by calculating the gradient intensity and direction of the image. Subsequently, non-maximum suppression (NMS) is used to refine the edges and remove false edge points. Then, a double thresholding algorithm is used to identify strong and weak edges, and the continuity of the edges is determined through connectivity analysis, thereby generating an accurate region contour set, providing precise geometric shape data for subsequent 3D reconstruction. After obtaining the region contour sets from different viewpoints, feature matching is performed on these images from different viewpoints according to common features between the images (such as corners, edges, etc.), and the spatial geometric relationship of the images is established by calculating the relative positional relationship between the images. Subsequently, based on the principle of triangulation, the actual coordinates of each point in space are calculated using the depth information of the image, thus mapping each contour point from the two-dimensional image to three-dimensional space, constructing a three-dimensional contour of the construction area. This three-dimensional contour not only describes the two-dimensional planar shape of the area but also provides height information, forming complete three-dimensional geometric data for subsequent construction planning and analysis. Then, the generated three-dimensional contour is used as a fitting framework, and point cloud fitting techniques (such as least squares method, RANSAC algorithm, etc.) are used to initially fit the area's point cloud set. The fitting accuracy between the point cloud data and the three-dimensional contour is calculated, and optimization adjustments are made based on the fitting accuracy to reduce the error between the point cloud data and the three-dimensional contour, thereby aligning the actually collected point cloud data with the three-dimensional contour to obtain a point cloud fitting result that best matches the actual construction environment. Finally, the optimal point cloud fitting result is combined with the three-dimensional contour and construction environment data to construct a three-dimensional construction twin model. This model realistically and accurately reflects the spatial characteristics of the construction area, providing detailed virtual environment support for simulation, planning, and monitoring during the construction process.

[0020] Furthermore, point cloud fitting is performed on the point cloud set of the region based on the three-dimensional contour of the region to obtain the optimal point cloud fitting result, including:

[0021] Using the three-dimensional contour of the region as a frame, a random fitting is performed on the point cloud set of the region to obtain a first point cloud fitting result, and a first fitting accuracy is calculated, wherein the fitting accuracy is the ratio of the number of point clouds falling within the three-dimensional contour of the region to the total number of point clouds in the point cloud set of the region; again using the three-dimensional contour of the region as a frame, a random fitting is performed on the point cloud set of the region to obtain a second point cloud fitting result and a second fitting accuracy; the random fitting of point clouds continues until a preset number of fittings is reached, and the point cloud fitting result corresponding to the maximum fitting accuracy is output as the optimal point cloud fitting result.

[0022] Optionally, during point cloud fitting, the point cloud set is first randomly fitted using the region's 3D contour as a framework. In this process, the region's 3D contour is defined as a constraint for point cloud fitting, and a subset is randomly selected from the region's point cloud set. The least squares method is used to minimize the error, thus fitting the relationship between the points in this subset and the 3D contour, obtaining the first point cloud fitting result. For the first point cloud fitting result, the first fitting accuracy is calculated by counting the number of points falling within the region's 3D contour and dividing this number by the total number of points in the region's point cloud set. Subsequently, based on the region's 3D contour, the point cloud set is randomly fitted again to obtain the second point cloud fitting result. This process is similar to the first, still constraining and optimizing the point cloud set within the region's 3D contour framework to obtain the second fit. For the second fitting result, the second fitting accuracy is calculated, again by counting the ratio of the number of points falling within the region's 3D contour to the total number of points. This process is repeated until a preset number of fittings is reached, ensuring that the fitting process sufficiently covers all possibilities in the point cloud data. After completing all random fittings, the fitting accuracy of each fitting is compared, and the fitting result corresponding to the highest fitting accuracy is selected as the final optimal point cloud fitting result. This result has the highest accuracy and can most accurately reflect the matching degree between the three-dimensional contour of the region and the point cloud data, providing accurate basic data for the subsequent construction of the three-dimensional construction twin model.

[0023] Based on the preset number of anchor points, the anchor points are randomly laid out within the cofferdam bridge construction area to generate several anchor point distributions.

[0024] In one embodiment, the predetermined number of anchor points is first determined based on design requirements or construction plans. Then, within the cofferdam bridge construction area, several locations are randomly selected for anchor point placement. This placement process does not rely on a fixed pattern or order; instead, the anchor points are randomly distributed throughout the construction area. The location of each anchor point is freely chosen while meeting certain constraints, such as avoiding overlap with existing structures or ensuring that each anchor point covers critical areas. Finally, through this random placement, multiple different anchor point distributions are generated, providing diverse references for subsequent simulation analysis and optimization.

[0025] Using the aforementioned three-dimensional construction twin model, external load simulation and construction simulation are performed based on the distribution of the aforementioned anchor points, and the fitness of several distributions is evaluated and determined based on the simulation results.

[0026] In one embodiment, after obtaining several anchor point distributions, these distributions are synchronized to a constructed 3D construction twin model. This 3D construction twin model can accurately reproduce the geometry and physical characteristics of the construction area. Subsequently, in the 3D construction twin model, external load simulation and construction simulation are performed on each anchor point distribution. External load simulation simulates various external forces (such as water flow velocity and load intensity) that may be applied to the anchor points during construction, calculating the bearing capacity and stability coefficient of each anchor point to evaluate its adaptability to loads in the actual environment. Construction simulation simulates the construction complexity and duration that the anchor point distribution would have to withstand during actual construction, evaluating the reliability of the anchor points during construction. Finally, based on the results of these two simulations, the distribution adaptability of each anchor point distribution is evaluated, i.e., how each layout performs in the face of loads and construction processes. A layout with higher distribution adaptability means that the anchor point distribution can better withstand external pressure and changes during construction, exhibiting higher construction safety and reliability.

[0027] Furthermore, using the aforementioned three-dimensional construction twin model, external load simulation and construction simulation are performed based on the distribution of the plurality of anchor points. Based on the simulation results, the fitness of several distributions is evaluated and determined, including:

[0028] A first anchor point distribution is randomly selected from the plurality of anchor point distributions; using the three-dimensional construction twin model, multiple external load simulations are performed based on external load factors and the first anchor point distribution to evaluate and determine the first bearing capacity and the first bearing stability coefficient; using the three-dimensional construction twin model, construction simulations are performed based on the first anchor point distribution to obtain the first construction complexity, the first construction time, and the first construction resource consumption; based on the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction time, and the first construction resource consumption, the first distribution fitness is evaluated and determined, and several distribution fitnesss of several anchor point distributions are obtained by sequential analysis.

[0029] Preferably, firstly, from the generated anchor point distributions, one anchor point distribution is randomly selected as the first anchor point distribution, which serves as the initial layout for subsequent simulation analysis. Then, using a three-dimensional construction twin model, multiple external load simulations are performed based on the selected first anchor point distribution. These external loads include water flow velocity, load intensity, etc. By simulating these external loads, the response of each anchor point under different loads can be calculated, quantifying the first bearing capacity and the first bearing stability coefficient. The first bearing capacity reflects the maximum force that the anchor point can withstand, and the first bearing stability coefficient reflects the stability of the anchor point under varying loads. Subsequently, using a 3D construction twin model, construction simulation is conducted based on the selected first anchor point distribution. This process simulates tasks such as anchor point installation, structural reinforcement, material handling, and equipment deployment, determining the first construction complexity, first construction duration, and first construction resource consumption for the first anchor point distribution. The first construction complexity reflects the technical difficulty and operational complexity of the entire construction process; the first construction duration refers to the time required to complete the entire construction process; and the first construction resource consumption reflects the total consumption of various resources, including the use of materials (such as building materials), equipment (such as machinery and transportation vehicles), and labor (such as the number of workers and working hours). Then, the simulated first bearing capacity, first bearing stability coefficient, first construction complexity, first construction duration, and first construction resource consumption are comprehensively evaluated, and a weighted average is used to obtain the first distribution fitness. A higher first distribution fitness indicates better performance in terms of bearing capacity, construction efficiency, and resource consumption. Finally, the remaining anchor point distributions are analyzed using the same method to obtain the fitness of each distribution. This helps to find the most suitable layout among various possible anchor point distributions, ensuring the safety, efficiency, and optimal use of resources during construction.

[0030] Furthermore, based on the aforementioned external load factors and the distribution of the first anchoring points, multiple external load simulations are performed to evaluate and determine the first bearing capacity and the first bearing stability coefficient, including:

[0031] External load factors are determined by analyzing external load records within a preset historical time zone; multiple external load simulations are performed based on the external load factors and the distribution of the first anchor points to determine multiple simulated bearing capacities, and the first bearing capacity is obtained by calculating the average value; bearing fluctuation is calculated on the multiple simulated bearing capacities, and the bearing fluctuation is subtracted from 1 to obtain the first bearing stability coefficient, wherein the bearing fluctuation is the ratio of the standard deviation of the multiple simulated bearing capacities to the first bearing capacity.

[0032] Optionally, during external load simulation, external load records within a preset historical time zone are extracted from the external load log. Monitoring data for preset external load indices, such as water flow velocity, water flow rate, load intensity (weight of construction equipment, weight of poured concrete, etc.), and wind speed, are then extracted from these records. The average of these preset external load monitoring data is calculated to obtain the external load factor, which includes the external load index and its corresponding value. Subsequently, the determined external load factor is applied to the distribution of the first anchorage point. Multiple external load simulations are performed using a three-dimensional construction twin model. Each simulation combines the external load factor with the anchorage point distribution, assessing the bearing capacity of each anchorage point based on the load action, generating multiple simulated bearing capacity results. By averaging these multiple simulated bearing capacity results, the first bearing capacity is obtained—the average bearing capacity of the anchorage point distribution under multiple load conditions. Next, the bearing capacity results of all simulations are statistically analyzed, and the standard deviation of the bearing capacity is calculated. The standard deviation reflects the degree of fluctuation of the bearing capacity under different load simulation conditions. The standard deviation is then divided by the first bearing capacity to obtain the bearing capacity fluctuation. The larger the bearing capacity fluctuation, the worse the stability of the anchor point under load changes. Then, this bearing capacity fluctuation is subtracted from 1 to obtain the first bearing capacity stability coefficient. This first bearing capacity stability coefficient reflects the adaptability and stability of the anchor point distribution to changes in external loads. The higher the stability coefficient, the stronger the distribution's ability to withstand load changes and the higher its stability. In summary, through the above process, the performance of each anchor point distribution under different external load factors can be comprehensively evaluated, including its bearing capacity, stability, and adaptability to environmental changes. This data helps determine the optimal anchor point layout, thereby ensuring the safety and efficiency of construction.

[0033] Furthermore, the first distribution fitness is determined based on the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction duration, and the first construction resource consumption assessment, including:

[0034] Based on the coefficient of variation method, weights are set for bearing capacity, bearing stability coefficient, construction complexity, construction time and construction resource consumption to obtain the weight ratio; based on the weight ratio, the first distribution fitness is obtained by weighting the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction time and the first construction resource consumption.

[0035] Optionally, after obtaining the first bearing capacity, first bearing stability coefficient, first construction complexity, first construction time, and first construction resource consumption, the relative importance of each indicator is determined using the coefficient of variation method. Specifically, for each indicator, the standard deviation of the indicator is divided by its mean to obtain the coefficient of variation. A higher coefficient of variation indicates greater volatility and may have a greater impact on the overall fitness in the assessment. Subsequently, the coefficient of variation of each indicator is normalized by dividing the coefficient of variation of each indicator by the sum of the coefficients of variation to calculate the weight of each indicator. A larger weight indicates a stronger influence of the indicator in the assessment. Then, based on the calculated weight, a weighted average method is used to calculate the first bearing capacity, first bearing stability coefficient, first construction complexity, first construction time, and first construction resource consumption to obtain the first distribution fitness. This first distribution fitness reflects the comprehensive adaptability of the first anchoring point distribution under all considered factors. A higher fitness value indicates that the anchoring point distribution performs better in terms of bearing capacity, construction stability, construction efficiency, and resource utilization, and is more suitable for implementation.

[0036] Based on the aforementioned distribution fitness, the anchor point distribution is iteratively optimized until convergence, and the optimal anchor point distribution is output for anchor point construction.

[0037] In one embodiment, during the iterative optimization of the anchor point distribution, multiple initial anchor point distributions are first used as initial solutions. These distributions are then divided based on their fitness values, identifying multiple optimal and multiple suboptimal solutions. The division of these optimal and suboptimal solutions is based on their fitness levels; optimal solutions perform better, while suboptimal solutions perform worse. Subsequently, suboptimal solutions are clustered around optimal solutions, adjusting them to regions closer to the optimal solutions to improve the overall fitness of the anchor point distribution. Following this, based on the clustering results, iterative optimization of the anchor point distribution is performed according to a preset adjustment step size. After each round of optimization, the anchor point layout is re-evaluated based on fitness until the fitness change reaches the convergence criterion, meaning the optimization process no longer significantly improves the results. Once the convergence condition is met, the optimal anchor point distribution with the best fitness is output as the final construction scheme for actual construction, thereby improving construction accuracy and efficiency, reducing risks, ensuring the smooth progress of the construction process, and guaranteeing the quality of the final project.

[0038] Furthermore, based on the aforementioned distribution fitness, the anchor point distribution is iteratively optimized until convergence, outputting the optimal anchor point distribution, including:

[0039] Using the anchor point distribution as the initial solution, based on the fitness of the distributions, the anchor point distributions are divided to determine K excellent solutions and P inferior solutions, where P is Q times K and Q is an integer greater than 10. The P inferior solutions are clustered around the K excellent solutions to obtain K solution sets. The anchor point distribution is iteratively optimized based on the K solution sets until convergence, and the optimal anchor point distribution is output.

[0040] Preferably, firstly, all anchor point distributions are used as initial solutions, and each distribution is evaluated based on its fitness value. Distributions with higher fitness are considered excellent solutions, and those with lower fitness are considered inferior solutions. This divides the anchor point distributions into K excellent solutions and P inferior solutions, where P is a multiple of K (Q is an integer greater than 10). This division ensures that a sufficient number of inferior solutions participate in optimization while maintaining the quality of excellent solutions. Subsequently, clustering is performed on the P inferior solutions centered around the K excellent solutions. The purpose of clustering is to adjust inferior solutions to their neighborhoods within the excellent solutions, thereby improving the overall fitness. During clustering, Euclidean distance is used to calculate the distance between inferior and excellent solutions, and inferior solutions are assigned to the nearest excellent solution, thus forming K solution sets. Subsequently, based on the obtained K solution sets, iterative optimization of the anchor point distribution is performed. In each round of optimization, a preset adjustment step size is used to adjust the layout position of inferior solutions, allowing them to gradually approach the optimal position under the influence of superior solutions. During the iteration process, superior solutions are used as reference points for adjustment and optimization to reduce errors and unreasonable layouts in the anchor point distribution. This optimization process continues until the fitness change is very small, reaching the preset convergence criterion. At the end of the iterative optimization, the anchor point distribution with the highest fitness is selected as the optimal anchor point distribution and used as the final construction scheme to ensure that stability, bearing capacity, and construction efficiency are maximized during construction.

[0041] Furthermore, based on the K solution sets, iterative optimization of the anchor point distribution is performed until convergence, outputting the optimal anchor point distribution, including:

[0042] Based on the K solution sets, within each solution set, with the optimal solution as the optimization direction, the inferior solutions within the solution set are optimized once according to a preset adjustment step size to obtain K updated solution sets. Based on the K updated solution sets, for each updated solution set, it is determined whether the distribution fitness of the inferior solution after the update is greater than or equal to the distribution fitness of the optimal solution. If so, the inferior solution is used to replace the optimal solution. Iteration optimization continues until a preset number of convergences is reached, and K current updated solution sets are output. The optimal solution of the optimal updated solution set is set as the optimal anchor point distribution, wherein the sum of the distribution fitness of the optimal updated solution set is the largest.

[0043] Optionally, in each solution set, using the optimal solution as the optimization direction, the inferior solutions within the solution set are optimized once according to a preset adjustment step size. During the optimization process, inferior solutions are gradually adjusted to the vicinity of the optimal solution. By adjusting their position and layout, the fitness of inferior solutions is improved as much as possible, tending towards the optimal solution. After optimization, the inferior solutions in each solution set will be improved to a certain extent, thus generating K updated solution sets. Subsequently, for the inferior solutions in each updated solution set, the simulation is repeated to obtain the distribution fitness of the updated inferior solutions. This new distribution fitness of inferior solutions is then compared with the fitness of the optimal solutions in the solution set. If the distribution fitness of the updated inferior solutions is higher or equal to that of the optimal solutions, the inferior solutions are used to replace the optimal solutions, that is, the inferior solutions are promoted to new optimal solutions. If the fitness of the updated inferior solutions is still lower than that of the optimal solutions, the optimal solutions are kept unchanged. Then, the optimization and replacement process continues until the preset number of convergences is reached. After each iteration, the fitness of the updated solution set is re-evaluated and compared. Each optimization will adjust the inferior solutions to further improve the overall fitness of the solution set. When the preset number of iterations is reached or the fitness change becomes very small, the optimization process is considered to have converged. At this point, the optimal updated solution set is selected from all updated solution sets. This optimal updated solution set is the one with the largest sum of fitness values ​​among all solution sets. Finally, the best solutions in the optimal updated solution set are extracted as the optimal anchor point distribution, providing the optimal layout scheme for actual construction and ensuring the stability, safety, and efficiency of the construction process.

[0044] Furthermore, for each updated solution set, it is determined whether the updated inferior solution exceeds the cofferdam bridge construction area. If it does, any qualified solution is randomly selected within the cofferdam bridge construction area to replace it.

[0045] Optionally, for each suboptimal solution in the updated solution set, check whether the updated suboptimal solution exceeds the boundary of the cofferdam bridge construction area. The boundary of the construction area is determined by the engineering design and construction specifications to ensure that the anchor point layout does not deviate from the workable range. If the updated suboptimal solution exceeds the boundary of the construction area, it means that the anchor point layout is infeasible and must be adjusted. In this case, a qualified solution, i.e., a layout scheme with high fitness within the construction area, will be randomly selected within the construction area to replace the suboptimal solution that exceeds the boundary. The replacement process ensures that the new suboptimal solution does not exceed the construction area, while maintaining the rationality and fitness of the anchor point layout as much as possible, thereby avoiding the use of invalid solutions and ensuring that the optimization process is always carried out within the legal construction area.

[0046] In summary, the digital twin-based anchorage point positioning method for cofferdam bridge construction provided in this application has the following technical effects:

[0047] First, drones are used to acquire multi-angle images and perform laser scanning of the cofferdam bridge construction area. A three-dimensional construction twin model is constructed by fusing the regional image set and regional point cloud set. Then, based on a preset number of anchor points, anchor points are randomly distributed within the cofferdam bridge construction area, generating several anchor point distributions. Next, using the three-dimensional construction twin model, external load simulation and construction simulation are performed based on the several anchor point distributions. The fitness of several distributions is evaluated and determined based on the simulation results. Finally, the anchor point distribution is iteratively optimized based on the fitness of the several distributions until convergence, and the optimal anchor point distribution is output for anchor point construction. This solves the technical problems in traditional cofferdam bridge construction where anchor point positioning relies on manual experience and static design drawings, resulting in large layout deviations and difficulty in adapting to dynamic construction disturbances. It achieves the technical effect of optimizing anchor point positions through digital twin models, improving construction accuracy, reducing construction errors, and thus improving construction efficiency and safety.

[0048] Example 2 is based on the same inventive concept as the digital twin-based anchorage point positioning method for cofferdam bridge construction in the previous examples, such as... Figure 2 As shown in the embodiment of this application, a digital twin-based anchorage point positioning system for cofferdam bridge construction is provided. The system includes:

[0049] Twin Modeling Module 11: Utilizes UAVs to acquire multi-angle images and perform laser scanning of the cofferdam bridge construction area, and constructs a three-dimensional construction twin model by fusing the regional image set and regional point cloud set; Anchor Point Layout Module 12: Based on a preset number of anchor points, randomly lays out anchor points within the cofferdam bridge construction area to generate several anchor point distributions; Model Simulation Module 13: Using the three-dimensional construction twin model, performs external load simulation and construction simulation based on the several anchor point distributions, and evaluates and determines the fitness of several distributions based on the simulation results; Anchor Point Construction Module 14: Iteratively optimizes the anchor point distribution based on the fitness of the several distributions until convergence, outputs the optimal anchor point distribution, and performs anchor point construction.

[0050] Furthermore, the twin modeling module 11 is used to perform the following operation steps:

[0051] The region image set is extracted using an edge detection algorithm to obtain a region contour set; a three-dimensional reconstruction is performed based on the region contour set according to a preset acquisition angle to output the region three-dimensional contour; the region point cloud set is fitted with the region point cloud set based on the region three-dimensional contour to obtain the optimal point cloud fitting result and construct a three-dimensional construction twin model.

[0052] Furthermore, the twin modeling module 11 is used to perform the following operation steps:

[0053] Using the three-dimensional contour of the region as a frame, a random fitting is performed on the point cloud set of the region to obtain a first point cloud fitting result, and a first fitting accuracy is calculated, wherein the fitting accuracy is the ratio of the number of point clouds falling within the three-dimensional contour of the region to the total number of point clouds in the point cloud set of the region; again using the three-dimensional contour of the region as a frame, a random fitting is performed on the point cloud set of the region to obtain a second point cloud fitting result and a second fitting accuracy; the random fitting of point clouds continues until a preset number of fittings is reached, and the point cloud fitting result corresponding to the maximum fitting accuracy is output as the optimal point cloud fitting result.

[0054] Furthermore, the model simulation module 13 is used to perform the following operation steps:

[0055] A first anchor point distribution is randomly selected from the plurality of anchor point distributions; using the three-dimensional construction twin model, multiple external load simulations are performed based on external load factors and the first anchor point distribution to evaluate and determine the first bearing capacity and the first bearing stability coefficient; using the three-dimensional construction twin model, construction simulations are performed based on the first anchor point distribution to obtain the first construction complexity, the first construction time, and the first construction resource consumption; based on the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction time, and the first construction resource consumption, the first distribution fitness is evaluated and determined, and several distribution fitnesss of several anchor point distributions are obtained by sequential analysis.

[0056] Furthermore, the model simulation module 13 is used to perform the following operation steps:

[0057] External load factors are determined by analyzing external load records within a preset historical time zone; multiple external load simulations are performed based on the external load factors and the distribution of the first anchor points to determine multiple simulated bearing capacities, and the first bearing capacity is obtained by calculating the average value; bearing fluctuation is calculated on the multiple simulated bearing capacities, and the bearing fluctuation is subtracted from 1 to obtain the first bearing stability coefficient, wherein the bearing fluctuation is the ratio of the standard deviation of the multiple simulated bearing capacities to the first bearing capacity.

[0058] Furthermore, the model simulation module 13 is used to perform the following operation steps:

[0059] Based on the coefficient of variation method, weights are set for bearing capacity, bearing stability coefficient, construction complexity, construction time and construction resource consumption to obtain the weight ratio; based on the weight ratio, the first distribution fitness is obtained by weighting the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction time and the first construction resource consumption.

[0060] Furthermore, the anchoring point construction module 14 is used to perform the following operation steps:

[0061] Using the anchor point distribution as the initial solution, based on the fitness of the distributions, the anchor point distributions are divided to determine K excellent solutions and P inferior solutions, where P is Q times K and Q is an integer greater than 10. The P inferior solutions are clustered around the K excellent solutions to obtain K solution sets. The anchor point distribution is iteratively optimized based on the K solution sets until convergence, and the optimal anchor point distribution is output.

[0062] Furthermore, the anchoring point construction module 14 is used to perform the following operation steps:

[0063] Based on the K solution sets, within each solution set, with the optimal solution as the optimization direction, the inferior solutions within the solution set are optimized once according to a preset adjustment step size to obtain K updated solution sets. Based on the K updated solution sets, for each updated solution set, it is determined whether the distribution fitness of the inferior solution after the update is greater than or equal to the distribution fitness of the optimal solution. If so, the inferior solution is used to replace the optimal solution. Iteration optimization continues until a preset number of convergences is reached, and K current updated solution sets are output. The optimal solution of the optimal updated solution set is set as the optimal anchor point distribution, wherein the sum of the distribution fitness of the optimal updated solution set is the largest.

[0064] Furthermore, the anchoring point construction module 14 is used to perform the following operation steps:

[0065] For each updated solution set, determine whether the updated inferior solution exceeds the cofferdam bridge construction area. If it does, randomly select any qualified solution within the cofferdam bridge construction area to replace it.

[0066] Through the foregoing detailed description of the anchorage point positioning method for cofferdam bridge construction based on digital twins, those skilled in the art can clearly understand the anchorage point positioning system for cofferdam bridge construction based on digital twins in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to in the method section.

[0067] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for locating anchorage points in cofferdam bridge construction based on digital twins, characterized in that the method... include: Using drones, multi-angle image acquisition and laser scanning were conducted on the construction area of ​​the cofferdam bridge. A three-dimensional construction twin model was constructed by fusing the regional image set and the regional point cloud set. Based on the preset number of anchor points, the anchor points are randomly laid out within the cofferdam bridge construction area to generate several anchor point distributions. Using the three-dimensional construction twin model, external load simulation and construction simulation are performed according to the distribution of the several anchor points, and the fitness of several distributions is evaluated and determined based on the simulation results; Based on the aforementioned distribution fitness, the anchor point distribution is iteratively optimized until convergence, and the optimal anchor point distribution is output for anchor point construction. Using the aforementioned three-dimensional construction twin model, external load simulation and construction simulation are performed based on the distribution of the aforementioned anchorage points. Based on the simulation results, the fitness of several distributions is evaluated and determined, including: The first anchor point distribution is randomly selected from the plurality of anchor point distributions; Using the aforementioned three-dimensional construction twin model, multiple external load simulations are performed based on external load factors and the distribution of the first anchorage points to evaluate and determine the first bearing capacity and the first bearing stability coefficient. Using the three-dimensional construction twin model, construction simulation is performed based on the distribution of the first anchoring points to obtain the first construction complexity, the first construction duration, and the first construction resource consumption. Based on the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction time, and the first construction resource consumption assessment, the first distribution fitness is determined, and several distribution fitnesss of several anchor point distributions are obtained by sequential analysis.

2. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 1, characterized in that, A 3D construction twin model is constructed by fusing regional image sets and regional point cloud sets, including: The region image set is obtained by extracting contours using an edge detection algorithm; According to the preset acquisition angle, perform three-dimensional reconstruction based on the region contour set, and output the three-dimensional contour of the region; Based on the three-dimensional contour of the region, the point cloud set of the region is fitted with point cloud to obtain the optimal point cloud fitting result and construct a three-dimensional construction twin model.

3. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 2, characterized in that, Based on the three-dimensional contour of the region, point cloud fitting is performed on the point cloud set of the region to obtain the optimal point cloud fitting result, including: Using the three-dimensional contour of the region as a framework, a random fitting is performed on the point cloud set of the region to obtain a first point cloud fitting result, and the first fitting accuracy is calculated. The fitting accuracy is the ratio of the number of point clouds falling within the three-dimensional contour of the region to the total number of point clouds in the point cloud set of the region. Using the three-dimensional contour of the region as a framework again, the point cloud set of the region is randomly fitted to obtain a second point cloud fitting result and a second fitting accuracy. Continue random point cloud fitting until the preset number of fitting attempts is reached, and output the point cloud fitting result corresponding to the maximum fitting accuracy as the optimal point cloud fitting result.

4. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 1, characterized in that, Based on the aforementioned external load factors and the distribution of the first anchorage points, multiple external load simulations are performed to evaluate and determine the first bearing capacity and the first bearing stability coefficient, including: External load factors are determined by analyzing external load records within a preset historical time zone. Based on the external load factors and the distribution of the first anchor points, multiple external load simulations are performed to determine multiple simulated bearing capacities, and the first bearing capacity is calculated from the average value. The load-bearing fluctuation is calculated for the multiple simulated load-bearing capacities. The first load-bearing stability coefficient is obtained by subtracting the load-bearing fluctuation from 1. The load-bearing fluctuation is the ratio of the standard deviation of the multiple simulated load-bearing capacities to the first load-bearing capacities.

5. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 1, characterized in that, The first distribution fitness is determined based on the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction duration, and the first construction resource consumption assessment, including: Based on the coefficient of variation method, weights are set for bearing capacity, bearing stability coefficient, construction complexity, construction time and construction resource consumption, and the weight ratios are obtained. Based on the weighted proportions, the first distribution fitness is obtained by weighting the first bearing capacity, the first bearing stability coefficient, the first construction complexity, the first construction duration, and the first construction resource consumption.

6. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 1, characterized in that, Based on the aforementioned distribution fitness, the anchor point distribution is iteratively optimized until convergence, outputting the optimal anchor point distribution, including: Using the anchor point distribution as the initial solution, based on the fitness of the distribution, the anchor point distribution is divided to determine K excellent solutions and P inferior solutions, where P is Q times K and Q is an integer greater than 10. Cluster the P inferior solutions with the K best solutions as the center to obtain the K solution set; Based on the K solution sets, iterative optimization of the anchor point distribution is performed until convergence, and the optimal anchor point distribution is output.

7. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 6, characterized in that, Based on the K solution sets, iterative optimization of the anchor point distribution is performed until convergence, outputting the optimal anchor point distribution, including: Based on the K solution sets, within each solution set, with the optimal solution as the optimization direction, the inferior solutions in the solution set are optimized once according to a preset adjustment step size to obtain K updated solution sets; Based on the K updated solution sets, for each updated solution set, determine whether the distribution fitness of the inferior solution after the update is greater than or equal to the distribution fitness of the superior solution. If so, replace the superior solution with the inferior solution. Continue iterative optimization until the preset number of convergences is reached, output K current updated solution sets, and set the optimal solution set as the optimal anchor point distribution, wherein the sum of the fitness of the optimal updated solution set distribution is the largest.

8. The method for locating anchorage points in cofferdam bridge construction based on digital twins according to claim 7, characterized in that, For each updated solution set, determine whether the updated inferior solution exceeds the cofferdam bridge construction area. If it does, randomly select any qualified solution within the cofferdam bridge construction area to replace it.

9. A digital twin-based anchorage point positioning system for cofferdam bridge construction, characterized in that, The system for implementing the digital twin-based anchorage point positioning method for cofferdam bridge construction according to any one of claims 1-8 includes: Twin modeling module: Use drones to collect multi-angle images and perform laser scanning on the construction area of ​​the cofferdam and bridge, and construct a three-dimensional construction twin model by fusing the regional image set and regional point cloud set; Anchor point layout module: Based on the preset number of anchor points, the anchor points are randomly laid out within the cofferdam bridge construction area to generate several anchor point distributions. Model simulation module: Using the three-dimensional construction twin model, external load simulation and construction simulation are performed according to the distribution of the several anchor points, and the fitness of several distributions is evaluated and determined based on the simulation results; Anchor point construction module: Based on the aforementioned distribution fitness, the anchor point distribution is iteratively optimized until convergence, and the optimal anchor point distribution is output for anchor point construction.

Images