Construction progress nonlinear dynamic monitoring method and system based on topological entropy iteration

Abstract

The invention discloses a construction progress nonlinear dynamic monitoring method and system based on topological entropy iteration.The method comprises the steps that firstly, capital construction site images, three-dimensional point cloud and environment data are collected to generate a multi-source comprehensive data set, then according to construction area, task and time division, an initial dynamic topological structure model is generated, an initial topological entropy value is calculated, and the initial dynamic topological structure model is obtained; updating the model and a topological entropy value, identifying a progress abnormal region by using an anomaly detection algorithm, analyzing and determining a key path and a node in combination with the key path, and quantifying an influence range of anomaly on overall construction; according to the method, a topological entropy iterative algorithm is introduced to combine multi-source data with a dynamic topological model, and a topological entropy value is iteratively updated in real time through distributed calculation, so that the nonlinear change characteristic of the construction progress is efficiently captured, and meanwhile, the topological complexity of a dynamic network structure in a complex construction scene is quantified; the key path and the construction bottleneck are accurately identified, the influence of abnormity on the whole construction plan is effectively evaluated, and the potential risk of construction period delay is reduced.

Description

Technical Field

[0001] The present invention relates to construction dynamic monitoring means, belonging to the field of infrastructure site monitoring, and particularly relates to a non-linear dynamic monitoring method and system for construction progress based on topological entropy iteration. Background Technique

[0002] With the development of digital and intelligent technologies, infrastructure site management is gradually moving towards informatization and automation. As an important part of infrastructure project management, construction progress monitoring can effectively reflect the plan execution of construction activities and the resource allocation efficiency. However, in the traditional construction management system, the monitoring and analysis of construction progress mainly rely on manual observation and static monitoring of fixed cameras, and combined with manual patrol and log recording methods to analyze construction progress. Although this method can provide surface information of construction activities to a certain extent, its intelligence level is low and it is difficult to meet the real-time progress monitoring requirements in dynamic and complex construction scenarios. On the one hand, fixed monitoring devices can only cover a limited monitoring range and it is difficult to obtain comprehensive dynamic data of the construction area. On the other hand, manual analysis has subjective judgment deviations and low efficiency, and cannot meet the requirements of large-scale infrastructure projects for real-time and accuracy.

[0003] Currently, some monitoring technologies based on drones or the Internet of Things are gradually applied to the infrastructure field to assist construction monitoring by collecting multi-source data on site. However, the following problems are exposed in the actual application of the existing technologies: First, the fusion processing ability of multi-source data is insufficient and it is impossible to efficiently construct a dynamic model of the site. Second, the existing monitoring technologies lack the ability to deeply model the non-linear dynamic characteristics of construction progress and it is difficult to accurately describe the complex evolution process of construction activities. In addition, the anomaly detection method based on empirical rules is difficult to adapt to the changing construction environment, resulting in low accuracy and reliability of progress anomaly detection, further restricting the intelligence level of construction management, and there are significant deficiencies in the real-time nature of construction progress monitoring, the accuracy of dynamic modeling, and the intelligence of anomaly detection, directly affecting the management efficiency and construction quality of infrastructure projects. Summary of the Invention

[0004] The purpose of the present invention is to overcome the above-mentioned defects and problems in the prior art and provide an accurate and efficient non-linear dynamic monitoring method and system for construction progress based on topological entropy iteration.

[0005] To achieve the above purpose, the technical solution of the present invention is: A non-linear dynamic monitoring method for construction progress based on topological entropy iteration, including:

[0006] S1. Collect image data, three-dimensional point cloud data and environmental monitoring data of the infrastructure site, generate a multi-source comprehensive data set, and perform standardized preprocessing to obtain a standardized multi-source comprehensive data set;

[0007] S2. Divide the standardized multi-source integrated dataset into multiple sub-datasets according to the construction area, construction tasks, and construction time, and generate an initial dynamic topological structure model of the infrastructure site through a triangular mesh division algorithm, based on the sub-datasets and combining the spatial location relationship between the construction area division and construction tasks.

[0008] S3. Based on the distributed computing architecture, calculate the topological entropy values of each construction area in the initial dynamic topological structure model at the initial time point respectively, and aggregate them to generate the global initial topological entropy.

[0009] S4. According to the time progress of the construction tasks and the real-time changes of the environmental monitoring data, use the distributed iterative method to update the initial dynamic topological structure model, and perform dynamic iterative calculations on the topological entropy values of each construction area.

[0010] S5. Calculate the topological entropy values after iterative update based on the anomaly detection algorithm, and identify the construction areas where significant changes occur before and after the topological entropy values after iterative update, and identify the construction progress anomaly areas.

[0011] S6. Combine the critical path analysis method to determine the critical path and critical nodes of the construction tasks, and quantify the impact range of the progress anomaly changes in the construction areas on the overall construction tasks.

[0012] The specific steps of step S1 include:

[0013] S11. Collect the image data, three-dimensional point cloud data, and environmental monitoring data of the infrastructure site.

[0014] The image data includes: the construction area boundary of the infrastructure site, the object surface texture information, and the spatial information of the construction area; the three-dimensional point cloud data includes the three-dimensional spatial position and structure information of the construction area; the environmental monitoring data includes temperature, humidity, and wind speed parameters.

[0015] S12. Align and fuse the image data, three-dimensional point cloud data, and environmental monitoring data of the infrastructure site according to the time stamp t to generate a multi-source integrated dataset D containing spatial information, time information, and resource information, and its expression is as follows:

[0016] D = {(I t , P t , E t ) | t ∈ [1, T]};

[0017] Where: I t is the image data at time t, P t is the three-dimensional point cloud data at time t, and E t is the environmental monitoring data at time t;

[0018] S13. Denoise the image data I of the infrastructure site based on a Gaussian filter t and perform distortion correction based on the camera internal parameter matrix K and the distortion parameter vector (k1, k2, p1, p2) to obtain the corrected image data I″ t ;

[0019] S14. Perform density smoothing on the 3D point cloud data P of the infrastructure site t ; adjust the density of the 3D point cloud data based on a filtering algorithm for nearest neighbor search, perform 3D coordinate alignment on the smoothed 3D point cloud data, and perform spatial alignment through the rigid body transformation matrix T = [R|t] to obtain the aligned 3D point cloud data P″ t ;

[0020] S15. Remove outliers from the environmental monitoring data E of the infrastructure site t and replace the data outside the range with linear interpolation based on the set upper and lower threshold values [L min , L max to obtain the environmental monitoring data E″ after outlier removal t ;

[0021] S16. Standardize the image data I″ t , the 3D point cloud data P″ t , and the environmental monitoring data E″ t , calculate the normalization values respectively, and integrate them into a standardized multi-source comprehensive dataset in a unified format. Its expression is as follows:

[0022] D′ = {(I″ t , P″ t , E″ t )|t ∈ [1, T]}.

[0023] The specific steps of step S2 include:

[0024] S21. Divide the standardized multi-source comprehensive dataset according to the construction area, construction task, and construction time, and classify the elements that meet the specific construction area, specific construction task, and specific construction time into a sub-dataset through the construction of a selection function Ψ. Its expression is as follows:

[0025] D′ j,k，t = Ψ(D′, R j , Q k , t) = {(I″ t，j，k , P″ t，j，k , E″ t，j，k )};

[0026] where: Ψ: D′ × R j × Q k × t → D′j，k，t is the partitioning operation function; I″ t，j，k At time t, in the construction area R j and for the construction task type Q k the corrected image data; P″ t，j，k is the aligned three-dimensional point cloud data under the same conditions as I″, E″ t，j，k is the environmental monitoring data after outlier removal and interpolation under the same conditions as I″ t，j，k is the environmental monitoring data after outlier removal and interpolation under the same conditions as I″ t，j，k ;

[0027] S22. Based on the three-dimensional point cloud data P″ t，j，k generate the triangular mesh partitioning result corresponding to time t, construction area R j , construction task type Q k , and use the triangular mesh partitioning algorithm to perform spatial dissection on P″ t，j，k to obtain a set of triangular elements, and its expression is as follows:

[0028]

[0029] where: T j，k，t is the set of triangular elements; is the three-point index forming the triangular element; U j，k，t is the total number of triangular elements generated by the partitioning; at time t, in the construction area R j , construction task type Q k the μth triangular element generated, corresponding to three point cloud point indices μ1, μ2, μ3; U j，k，t is the number of generated triangular elements; at time t, in the construction area R j , construction task type Q k the th L j，k，t the number of points contained in the three-dimensional point cloud data set; x, y, z are the three-dimensional point cloud coordinates;

[0030] S23. Define the initial regional connection relationship based on the set of triangular elements T j，k，t and construct the initial regional connection matrix A j，k，t , and its expression is as follows:

[0031]

[0032] where: a u，v is an element of the initial regional connection matrix; are any two triangular elements;

[0033] S24. Use the environmental monitoring data E″ t，j，k to process the initial regional connection matrix Aj，k，t Perform weighting to obtain the weighted initial regional topology matrix W j，k，t , and its expression is as follows:

[0034]

[0035] Where: W u，v is an element of the weighted initial regional topology matrix;

[0036] S25. Combine the weighted initial regional topology matrices, corresponding sub-datasets, and triangular element sets obtained at different times t, different construction regions R j , and different construction task types Q k according to the time series and spatial distribution to obtain the initial dynamic topology structure model, and its expression is as follows:

[0037] M = {(D′ j，k，t , T j，k，t , W j，k，t )|j ∈ [1, J], k ∈ [1, K], t ∈ [1, T]};

[0038] Where: M is the initial dynamic topology structure model.

[0039] The specific steps of step S3 include:

[0040] S31. Allocate tasks to the weighted initial regional topology matrices of each construction region and each construction task type based on the initial dynamic topology structure model according to distributed computing nodes;

[0041] S32. On each distributed computing node, extract the local weighted adjacency relationship from the weighted initial regional topology matrix, and use the normalization method to calculate the transfer probability matrix to obtain the weighted transfer probability;

[0042] S33. Define a regional optimization model for calculating topological entropy based on the weighted transfer probability, calculate the topological entropy values returned by each distributed computing node, and aggregate them to generate the global initial topological entropy, and its expression is as follows:

[0043]

[0044] Where: is the global initial topological entropy, is the topological entropy value, ψ j，k (p u，v ) = p u,v ·ξ(E′ t，j，k ) is the weighted result of the transfer probability between nodes.

[0045] The specific steps of step S4 include:

[0046] S41. Based on the initial dynamic topology structure model and the topological entropy value, allocate the weighted initial regional topology matrix W corresponding to each moment t in the initial dynamic topology structure model, j，k，t the environmental detection data E″, t，j，k and the topological entropy value to the distributed computing nodes to generate a dynamic topology update task set T j，k，t ;

[0047] S42. On each distributed computing node, according to the time progress of the construction task and the real-time changes of the environmental monitoring data, use the initial dynamic topology structure model (T j，k，t , W j，k，t ) and adopt a dynamic weight adjustment function to update the weighted initial regional topology matrix;

[0048] S43. Based on the updated weighted initial regional topology matrix W′ j，k，t , recalculate the topological entropy value H′ of each construction area and construction task at moment t according to the topological entropy value calculation method j，k，t ;

[0049] S44. Combine the changes in the topological entropy value H′ j，k，t over time series to calculate the dynamic change rate ΔH of the construction area and the construction task j，k，t , and its expression is as follows:

[0050]

[0051] where: Δt is the adjacent time step.

[0052] The specific steps of step S5 include:

[0053] S51. Based on the dynamic change rate ΔH j，k，t and combined with the statistical characteristics of the dynamic change rate, define an anomaly detection threshold θ j，k , and its expression is as follows:

[0054] θ j，k = μ j，k + λ·σ j,k ;

[0055] where: μ j，k is the mean value of the dynamic change rate under the construction area R j , construction task type Q k , σ j，k is the standard deviation of the dynamic change rate, and λ is an adjustment parameter;

[0056] S52. Compare the dynamic change rate ΔH j，k，t with the anomaly detection threshold θ j，k ; if |ΔH j，k，t|>θ j，k indicates the construction area R j and the construction task type Q k has an abnormal dynamic change at time t, then record the abnormal time point and the abnormal change area A = {(j, k, t)||ΔH j，k,t |>θ j，k};

[0057] S53. Calculate the deviation degree of the dynamic change rate of the abnormal change area, and classify the construction progress abnormal area based on the deviation degree.

[0058] The step S6 specifically includes:

[0059] S61. Based on the initial dynamic topology structure model (T j，k，t , W j，k，t ) and the updated weighted initial area topology matrix W′ j，k，t , construct the dynamic critical path network model CPM j，k of the construction task, and its expression is as follows:

[0060]

[0061] where: (p u , p v , w′ u，v ) is the dynamic weighted path from node p u to node p v in the construction task, and w′ u，v is the path weight;

[0062] S62. Based on the dynamic critical path network model, analyze the cumulative weights of all paths through the critical path algorithm, and determine the set of critical paths and nodes, and its expression is as follows:

[0063]

[0064] where: CP j，k is the set of critical paths under the construction area R j and the construction task type Q k . The path with the largest cumulative weight of the path is the critical path of the construction area R j and the construction task type Q k ;

[0065] S63. Based on the set of critical paths CP j，k , update the topology structure of the abnormal change area (j, k, t) ∈ A, conduct a propagation analysis on the impact of the critical path, and quantify the impact range of the progress abnormal change of the area on the overall construction task, and its expression is as follows:

[0066]

[0067] where: δCP j，k is the influence ratio of the construction progress abnormal area on the cumulative weight of the critical path.

[0068] A non - linear dynamic monitoring system for construction progress based on topological entropy iteration, which is applied to the above - mentioned method. The system includes:

[0069] A multi - source integrated dataset construction module, which is used to collect image data, three - dimensional point cloud data and environmental monitoring data of the infrastructure site, generate a multi - source integrated dataset, and perform standardized pre - processing to obtain a standardized multi - source integrated dataset;

[0070] An initial dynamic topological structure model construction module, which is used to divide the standardized multi - source integrated dataset into multiple sub - datasets according to construction areas, construction tasks and construction time, and generate an initial dynamic topological structure model of the infrastructure site through a triangular mesh division algorithm based on the spatial position relationship between the sub - datasets, construction area division and construction tasks;

[0071] A global initial topological entropy calculation module, which is used to calculate the topological entropy values of each construction area in the initial dynamic topological structure model at the initial time point respectively based on a distributed computing architecture, and aggregate them to generate a global initial topological entropy;

[0072] A distributed iteration module, which is used to update the initial dynamic topological structure model by using a distributed iteration method according to the time progress of construction tasks and the real - time changes of environmental monitoring data, and perform dynamic iterative calculation on the topological entropy values of each construction area;

[0073] An abnormal area identification module, which is used to calculate the topological entropy values after iterative update based on an anomaly detection algorithm, and identify the construction areas where significant changes occur before and after the topological entropy values after iterative update, and identify the construction progress abnormal areas;

[0074] An anomaly quantification module, which is used to determine the critical path and critical nodes of construction tasks by combining the critical path analysis method, and quantify the influence range of the abnormal progress change of the construction area on the overall construction task.

[0075] A non - linear dynamic monitoring device for construction progress based on topological entropy iteration, the device includes a processor and a memory;

[0076] The memory is used to store computer program code and transmit the computer program code to the processor;

[0077] The processor is used to execute the above - mentioned non - linear dynamic monitoring method for construction progress based on topological entropy iteration according to the instructions in the computer program code.

[0078] A computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed on a computer, the above-mentioned non-linear dynamic monitoring method for construction progress based on topological entropy iteration is realized.

[0079] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0080] In the non-linear dynamic monitoring method and system for construction progress based on topological entropy iteration of the present invention, the method first collects infrastructure site images, three-dimensional point clouds and environmental monitoring data to generate a multi-source comprehensive data set and perform standardized processing, and then divides it according to construction areas, tasks and time to generate an initial dynamic topological structure model, calculates the initial topological entropy value, and then dynamically updates the model and the topological entropy value according to construction progress and environmental changes, uses an anomaly detection algorithm to identify the progress anomaly area, and combines critical path analysis to determine the critical path and nodes, quantifies the impact range of the anomaly on the overall construction, and realizes the accurate monitoring and evaluation of infrastructure construction progress; in the application of this design, the topological entropy iteration algorithm is introduced to combine the multi-source data of the construction site with the dynamic topological model, and the topological entropy value is iteratively updated in real time through distributed computing, dynamically and efficiently capturing the non-linear change characteristics of the construction progress, while fully quantifying the topological complexity of the dynamic network structure in complex construction scenarios, accurately identifying the critical path and construction bottlenecks, and effectively evaluating the impact of anomalies on the overall construction plan, reducing the potential risk of construction period delay. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 is the method flow chart of the present invention.

[0082] Figure 2 is the system structure diagram of the present invention.

[0083] Figure 3 is the equipment structure diagram of the present invention.

[0084] In the figure: multi-source comprehensive data set construction module 1, initial dynamic topological structure model construction module 2, global initial topological entropy calculation module 3, distributed iteration module 4, anomaly area identification module 5, anomaly quantification module 6, processor 7, memory 8, computer program code 81. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0085] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0086] Example 1:

[0087] See Figure 1 , a non-linear dynamic monitoring method for construction progress based on topological entropy iteration, including:

[0088] S1. Collect the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site, generate a multi-source comprehensive dataset, and perform standardized preprocessing to obtain a standardized multi-source comprehensive dataset;

[0089] Further, the step S1 specifically includes:

[0090] S11. Collect the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site;

[0091] The image data includes: the boundary of the construction area at the infrastructure construction site, the surface texture information of the object, and the spatial information of the construction area; the 3D point cloud data includes the 3D spatial position and structural information of the construction area; the environmental monitoring data includes temperature, humidity, and wind speed parameters;

[0092] In this embodiment, an unmanned aerial vehicle remote sensing device is used to collect the image data of the infrastructure construction site; a lidar device is used to collect the 3D point cloud data; and a fixed sensor is used to collect the environmental monitoring data.

[0093] S12. Align and fuse the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site according to the time stamp t to generate a multi-source comprehensive dataset D containing spatial information, time information, and resource information, and its expression is as follows:

[0094] D = {(I t , P t , E t ) | t ∈ [1, T]};

[0095] Where: I t is the image data at time t, P t is the 3D point cloud data at time t, and E t is the environmental monitoring data at time t;

[0096] S13. Denoise the image data I t of the infrastructure construction site based on a Gaussian filter, and perform distortion correction based on the camera internal parameter matrix K and the distortion parameter vector (k1, k2, p1, p2) to obtain the corrected image data I″ t ;

[0097] S14. Perform density smoothing on the 3D point cloud data P t of the infrastructure construction site; adjust the density of the 3D point cloud data based on a filtering algorithm of nearest neighbor search, align the 3D coordinates of the smoothed 3D point cloud data, and perform spatial alignment through the rigid body transformation matrix T = [R|t] to obtain the aligned 3D point cloud data P″ t ;

[0098] S15. For the environmental monitoring data E of the infrastructure construction sitet Outlier rejection is performed, and based on the set upper and lower threshold values [L min , L max , the data outside the range is replaced with linear interpolation to obtain the post-rejection environmental monitoring data E″ t ;

[0099] S16. Standardize the image data I″ t , the three-dimensional point cloud data P″ t , and the environmental monitoring data E″ t , calculate the normalization values respectively, and integrate them into a standardized multi-source comprehensive data set in a unified format. The expression is as follows:

[0100] D′ = {(I″ t , P″ t , E″ t )|t ∈ [1, T]}.

[0101] S2. Divide the standardized multi-source comprehensive data set into multiple sub-data sets according to the construction area, construction task, and construction time, and through the triangular mesh division algorithm, generate the initial dynamic topological structure model of the infrastructure site based on the spatial position relationship between the sub-data sets combined with the construction area division and construction task;

[0102] The step S2 specifically includes:

[0103] S21. Divide the standardized multi-source comprehensive data set according to the construction area, construction task, and construction time. By constructing a selection function Ψ, the elements that meet the specific construction area, specific construction task, and specific construction time are classified into sub-data sets. The expression is as follows:

[0104] D′ j，k，t = Ψ(D′, R j , Q k , t) = {(I t，j，k , P″ t，j，k , E″ t，j，k )};

[0105] Where: Ψ: D′ × R j × Q k × t → D j，k,t is a division operation function that filters and maps D′ according to the spatial numbering rule of the construction area R j , the classification identifier of the construction task type Q k , and the sequence index of the moment t to ensure that the sub-data set contains both the image data I″ t，j，k , the three-dimensional point cloud data P″ t，j，k , and the environmental monitoring data E″ t，j，k ; I″ t，j，kis the corrected image data at time t in the construction area R j and under the construction task type Q k ; P″ t，j，k is the aligned three-dimensional point cloud data E″ under the same conditions as I" t，j，k ; t，j，k is the environmental monitoring data after outlier removal and interpolation under the same conditions as I″ t，j，k ;

[0106] S22. Generate the triangular mesh division result corresponding to time t, construction area R t，j，k and construction task type Q j using the triangular mesh division algorithm to perform spatial dissection on P″ k to obtain a set of triangular elements. Let the three-dimensional point cloud data include L t，j，k points, and its expression is as follows: j，k，t

[0107]

[0108] where: T j，k，t is the set of triangular elements; is the index of the three points that make up the triangular element; U j，k，t is the total number of triangular elements generated by the division. Each triangular element is used to describe the basic spatial topology unit at this moment, in this area, and under this task type; is the μ-th triangular element generated at time t, in construction area R j and construction task type Q k , corresponding to three point cloud point indices μ1, μ2, μ3; U j，k，t is the number of generated triangular elements; is the number of points contained in the j -th L k three-dimensional point cloud data set at time t, in construction area R and construction task type Q j，k，t ; X, y, Z are the three-dimensional point cloud coordinates;

[0109] S23. Define the initial regional connection relationship based on the set of triangular elements T j，k，t to construct the initial regional connection matrix A j，k，t , and its expression is as follows:

[0110]

[0111] For any two triangular elements and , if they share a common boundary, that is, two point coordinates are the same, then:

[0112] ​

[0113] Wherein: a u，v is an element of the initial regional connection matrix, used to quantify the spatial adjacency relationship at time t in the construction area R j and the construction task type Q k ; are any two triangular elements;

[0114] S24. Using the environmental monitoring data E″ t，j，k to weight the initial regional connection matrix A j，k，t to obtain the weighted initial regional topology matrix W j，k，t , and its expression is as follows:

[0115]

[0116] Wherein: W u，v is an element of the weighted initial regional topology matrix;

[0117] S25. Combining the weighted initial regional topology matrices obtained at different times t, different construction areas R j , different construction task types Q k with the corresponding sub-datasets and the set of triangular elements according to the time series and spatial distribution to obtain the initial dynamic topology structure model, and its expression is as follows:

[0118] M = {(D′ j，k，t , T j，k，t , W j,k，t )|j ∈ [1, J], k ∈ [1, K], t ∈ [1, T]};

[0119] Wherein: M is the initial dynamic topology structure model. In the initial dynamic topology structure model, M simultaneously contains the spatio-temporal topology structure information after construction area division, construction task type classification, and environmental data weighting processing, reflecting the dynamic spatial configuration of the infrastructure site progress in different scenarios.

[0120] S3. Based on the distributed computing architecture, calculate the topological entropy values of each construction area in the initial dynamic topology structure model at the initial time point respectively, and aggregate them to generate the global initial topological entropy;

[0121] The step S3 specifically includes:

[0122] S31. Based on the initial dynamic topology structure model, allocate tasks to the weighted initial regional topology matrices of each construction area and each construction task type according to the distributed computing nodes, so that the tasks are load-balanced among the computing nodes, and at the same time initialize the computing resources to support distributed computing;

[0123] S32. On each distributed computing node, extract the local weighted adjacency relationship from the weighted initial regional topology matrix, and use the normalization method to calculate the transfer probability matrix to obtain the weighted transfer probability, giving the dynamic weight adjustment of environmental factors to the node connection probability within the region. The weight is determined by the influence function of the regional environmental monitoring data to quantify the dynamic interaction between the environment and topological complexity;

[0124] S33. Based on the regional optimization model for calculating topological entropy defined by the weighted transfer probability, calculate the topological entropy value returned by each distributed computing node, and aggregate to generate the global initial topological entropy. Its expression is as follows:

[0125]

[0126] where: is the global initial topological entropy, is the topological entropy value, ψ j，k (p u,v ) = p u,v ·ξ(E′ t，j，k ) is the weighted result of the transfer probability between nodes. By introducing the dynamic weight of environmental monitoring data, the topological entropy not only reflects the complexity of the spatial structure but also the real-time impact of the environment on the construction dynamics.

[0127] S4. According to the time progress of the construction tasks and the real-time changes of the environmental monitoring data, use the distributed iterative method to update the initial dynamic topological structure model, and perform dynamic iterative calculation on the topological entropy values of each construction area;

[0128] The step S4 specifically includes:

[0129] S41. Based on the initial dynamic topological structure model and the topological entropy value, assign the weighted initial regional topology matrix W j，k，t , the environmental detection data E″ t，j，k and the topological entropy value corresponding to each moment t in the initial dynamic topological structure model to the distributed computing nodes to generate the dynamic topology update task set T j，k，t ;

[0130] S42. On each distributed computing node, according to the time progress of the construction tasks and the real-time changes of the environmental monitoring data, use the initial dynamic topological structure model (T j，k，t , W j，k，t ) and adopt the dynamic weight adjustment function to update the weighted initial regional topology matrix;

[0131] S43. Based on the updated weighted initial regional topology matrix W′ j，k，t , recalculate the topological entropy value H′ of each construction area and construction task at moment t according to the topological entropy value calculation methodj，k，t ;

[0132] S44. Combine the topological entropy value H′ at different time steps j，k，t to calculate the dynamic change rate ΔH of the construction area and construction tasks j，k，t . Its expression is as follows:

[0133]

[0134] where: Δt is the adjacent time step. ΔH j，k，t is used to characterize the non - linear dynamic change rate of the construction area and construction task type. By analyzing ΔH j，k，t , the key time points of dynamic changes and abnormal change areas can be identified.

[0135] S5. Calculate the topologically - entropy value after iterative update based on the anomaly - detection algorithm, and identify the construction areas where significant changes occur before and after the iterative update of the topologically - entropy value, and identify the abnormal construction - progress areas;

[0136] The step S5 specifically includes:

[0137] S51. Based on the dynamic change rate ΔH j，k，t and combined with the statistical characteristics of the dynamic change rate, define the anomaly - detection threshold θ j，k . Its expression is as follows:

[0138] θ j，k = μ j，k + λ·σ j,k

[0139] where: μ j，k is the mean value of the dynamic change rate under the construction area R j and construction task type Q k ; σ j，k is the standard deviation of the dynamic change rate; λ is an adjustment parameter used to control the sensitivity of anomaly detection;

[0140] S52. Compare the dynamic change rate ΔH j，k，t with the anomaly - detection threshold θ j，k . If |ΔH j，k，t | > θ j，k , it means that there is an abnormal dynamic change in the construction area R j and construction task type Q k at time t. Then record the abnormal time point and the abnormal change area A = {(j, k, t)||ΔH j，k，t | > θ j，k};

[0141] S53. Calculate the deviation degree of the dynamic change rate of the abnormal change area, and classify the construction progress abnormal area based on the deviation degree.

[0142] In this solution, the deviation degree of the dynamic change rate represents the significance intensity of the abnormal area and the time point, and is used to classify the abnormal area and generate an abnormal classification report. The higher the significance intensity of the area, the higher the priority, that is, the higher the degree of abnormality.

[0143] S6. Combine the critical path analysis method to determine the critical path and critical nodes of the construction tasks, and quantify the influence scope of the progress abnormal change in the construction area on the overall construction tasks.

[0144] The step S6 specifically includes:

[0145] S61. Based on the initial dynamic topology structure model (T j，k，t , W j，k，t ) and the updated weighted initial area topology matrix W′ j，k，t , construct the dynamic critical path network model CPM j，k of the construction tasks, and its expression is as follows:

[0146]

[0147] Where: (p u , p v , w′ u，v ) is the dynamic weighted path from node p u to node p v in the construction tasks; w′ u，v is the path weight, which reflects the intensity of the dependency relationship between tasks;

[0148] S62. Based on the dynamic critical path network model, analyze the cumulative weights of all paths through the critical path algorithm to determine the set of critical paths and nodes, and its expression is as follows:

[0149]

[0150] Where: CP j，k is the set of critical paths under the construction area R j and the construction task type Q k . The path with the largest cumulative weight of the paths is the critical path of the construction area R j and the construction task type Q k ;

[0151] S63. Based on the set of critical paths CP j，kUpdate the topological structure of the abnormal change area (j, k, t) ∈ A, conduct a propagation analysis on the impact on the critical path, and quantify the impact scope of the abnormal change in the progress of the area on the overall construction task. The expression is as follows:

[0152]

[0153] Among them: δCP j，k Is the influence ratio of the abnormal construction progress area on the cumulative weight of the critical path.

[0154] In this embodiment, at the construction site of a certain infrastructure project, the construction of the main building of a high-speed railway station in a certain city is in progress for track laying and main building construction. The construction site area of the project is about 80,000 square meters, involving multiple task areas. Among them, area A is in progress for steel structure installation, area B is for track laying, and area C is for electrical equipment laying. 6 drone acquisition devices, 4 lidars and 30 environmental sensors are deployed on site to collect image data, three-dimensional point cloud data and environmental monitoring data of the construction infrastructure site in real time.

[0155] At 10 am, the image data collected by the drone and the three-dimensional point cloud data of the lidar are preprocessed to construct a dynamic topological structure model. Combining the environmental data of the day (temperature 35°C, humidity 25%, wind speed 2.5 m / s), the system preliminarily calculates the initial topological entropy values of each construction area; at this time, the topological entropy value of area A is 0.75, that of area B is 0.92, and that of area C is 0.68, and the change trends are all within the normal range.

[0156] At 11:30 am, the environmental sensor detects that the temperature in area B rises to 38°C and the wind speed increases to 4.8 m / s. The system updates the weighted initial area topological matrix of area B in real time through the dynamic weight adjustment function, and recalculates the topological entropy value of this area. It is found that its topological entropy value drops sharply from 0.92 to 0.58 in a short time, and the change rate is -0.34. The system automatically triggers the anomaly detection algorithm to calculate that the deviation degree of this change rate is 2.5 times the standard deviation, exceeding the anomaly detection threshold, and determines that there is an abnormal dynamic change in the track laying task in area B.

[0157] After further analysis, it is found that the specific reason for the anomaly is that the sudden change in high temperature and wind speed causes the track material to expand and the laying accuracy to decrease, which in turn causes the mechanical equipment to operate unstably; the system generates this anomaly report record and records the time point of the anomaly (11:30), the regional location (area B), the task type (track laying), the change trend of the topological entropy value, and the anomaly influence scope. At the same time, the system automatically sends a warning message to the management platform at the construction site.

[0158] After receiving the warning information, the construction management personnel quickly took emergency measures, including dispatching additional cooling equipment to Area B, reducing the operating speed of the track laying equipment by 10%, and redeploying construction workers and equipment to correct the laying errors. Subsequently, by 13:00 pm, the topographical entropy value of Area B recalculated by the system gradually recovered to 0.88, and the change rate returned to the normal range, and the abnormal state was lifted.

[0159] At 15:00 pm on the same day, the system evaluated the global impact range of the anomaly based on the critical path analysis algorithm. The results showed that the impact ratio of this anomaly on the cumulative weight of the critical path was 5.2%. It was originally expected that it might cause a global project duration delay of about 1 day, but through timely adjustment, the delay was controlled within 15 minutes.

[0160] To verify the effectiveness of the method of the present invention, a detailed comparison was made with the traditional fixed monitoring and manual analysis methods. The comparison results are shown in the following table:

[0161]

[0162] The solution system was initially trained based on several groups of historical construction data. The data covered various complex working conditions such as high temperature, heavy rain, and low temperature, and the total cumulative sample volume was 2TB. After training, the accuracy rate of anomaly detection reached over 95%. In actual tests, the method of the present invention successfully identified 3 anomaly events during the operation of this project, and reduced the potential project duration delay by a total of 5 days through critical path analysis, fully demonstrating the reliability and practicality of the method.

[0163] Embodiment 2:

[0164] See Figure 2 , a non-linear dynamic monitoring system for construction progress based on topological entropy iteration. This system is applied to the method described in Embodiment 1. The system includes:

[0165] A multi-source comprehensive data set construction module 1, which is used to collect image data, three-dimensional point cloud data, and environmental monitoring data of the infrastructure site, generate a multi-source comprehensive data set, and perform standardized preprocessing to obtain a standardized multi-source comprehensive data set;

[0166] Further, the multi-source comprehensive data set construction module 1 is used to construct a multi-source comprehensive data set according to the following steps:

[0167] S11. Collect image data, three-dimensional point cloud data, and environmental monitoring data of the infrastructure site;

[0168] The image data includes: the construction area boundary of the infrastructure site, the object surface texture information, and the spatial information of the construction area; the three-dimensional point cloud data includes the three-dimensional spatial position and structural information of the construction area; the environmental monitoring data includes temperature, humidity, and wind speed parameters;

[0169] S12. Align and fuse the image data, 3D point cloud data, and environmental monitoring data at the infrastructure construction site according to the time stamp t to generate a multi-source comprehensive dataset D containing spatial information, time information, and resource information. The expression is as follows:

[0170] D = {(I t , P t , E t ) | t ∈ [1, T]};

[0171] Where: I t is the image data at time t, P t is the 3D point cloud data at time t, and E t is the environmental monitoring data at time t;

[0172] S13. Denoise the image data I t at the infrastructure construction site based on a Gaussian filter, and perform distortion correction based on the camera internal parameter matrix K and the distortion parameter vector (k1, k2, p1, p2) to obtain the corrected image data I″ t ;

[0173] S14. Perform density smoothing on the 3D point cloud data P t at the infrastructure construction site; adjust the density of the 3D point cloud data based on a filtering algorithm using nearest neighbor search, perform 3D coordinate alignment on the smoothed 3D point cloud data, and perform spatial alignment through the rigid body transformation matrix T = [R|t] to obtain the aligned 3D point cloud data P″ t ;

[0174] S15. Remove outliers from the environmental monitoring data E t at the infrastructure construction site, and replace the data outside the set upper and lower threshold values [L min , L max with linear interpolation to obtain the outlier-removed environmental monitoring data E″ t ;

[0175] S16. Standardize the image data I″ t , the 3D point cloud data P″ t , and the environmental monitoring data E″ t , calculate the normalization values respectively, and integrate them into a standardized multi-source comprehensive dataset in a unified format. The expression is as follows:

[0176] D′ = {(I″ t , P″ t , E″ t ) | t ∈ [1, T]}.

[0177] The initial dynamic topology structure model construction module 2 is used to divide the standardized multi-source integrated dataset into multiple sub-datasets according to the construction area, construction task, and construction time, and generate the initial dynamic topology structure model of the infrastructure site through the triangular mesh division algorithm based on the sub-datasets in combination with the spatial position relationship between the construction area division and the construction task;

[0178] Further, the initial dynamic topology structure model construction module 2 is used to construct the initial dynamic topology structure model according to the following steps:

[0179] S21. Divide the standardized multi-source integrated dataset according to the construction area, construction task, and construction time, and classify the elements that meet the specific construction area, specific construction task, and specific construction time into sub-datasets through the construction of the selection function Ψ. Its expression is as follows:

[0180] D′ j,k，t = Ψ(D′, R j , Q k , t) = {(I″ t，j，k , P″ t，j，k , E″ t，j，k )}

[0181] Where: Ψ: D′×R j ×Q k ×t → D' j，k，t is the division operation function; I″ t，j，k is the corrected image data at time t, within the construction area R j , and under the construction task type Q k ; P″ t，j，k is the aligned 3D point cloud data under the same conditions as I″ t，j，k , and E″ t，j，k is the environmental monitoring data after outlier removal and interpolation under the same conditions as I″ t，j，k ;

[0182] S22. Generate the triangular mesh division result corresponding to time t, construction area R t，j，k , and construction task type Q j based on the 3D point cloud data P″ k , and use the triangular mesh division algorithm to perform spatial subdivision on P″ t，j，k to obtain the triangular element set. Its expression is as follows:

[0183]

[0184] Where: T j，k，t is the triangular element set; is the three-point index constituting the triangular element; U j，k，t is the total number of triangular elements generated by the division; For the μ-th triangular element generated at time t in the construction area R j and the construction task type Q k it corresponds to three point cloud point indices μ1, μ2, μ3; U j，k，t is the number of triangular elements generated; For time t, construction area R j and the construction task type Q k the th L j，k，t the number of points contained in the three-dimensional point cloud dataset; X, y, z are the three-dimensional point cloud coordinates;

[0185] S23. Based on the triangular element set T j，k，t define the initial regional connection relationship and construct the initial regional connection matrix A j，k，t , and its expression is as follows:

[0186]

[0187] where: a u，v is an element of the initial regional connection matrix; are any two triangular elements;

[0188] S24. Use the environmental monitoring data E″ t，j，k to weight the initial regional connection matrix A j，k，t to obtain the weighted initial regional topology matrix W j，k，t , and its expression is as follows:

[0189]

[0190] where: W u，v is an element of the weighted initial regional topology matrix;

[0191] S25. Combine the weighted initial regional topology matrices obtained at different times t, different construction areas R j , and different construction task types Q k along with the corresponding sub-datasets and triangular element sets according to the time series and spatial distribution to obtain the initial dynamic topology structure model, and its expression is as follows:

[0192] M = {(D′ j，k，t , T j，k，t , W j，k，t )|j ∈ [1, J], k ∈ [1, K], t ∈ [1, T]};

[0193] where: M is the initial dynamic topology structure model.

[0194] The global initial topological entropy calculation module 3 is used to calculate the topological entropy values of each construction area in the initial dynamic topological structure model at the initial time point based on the distributed computing architecture, and aggregate them to generate the global initial topological entropy;

[0195] Further, the global initial topological entropy calculation module 3 is used to calculate the global initial topological entropy according to the following steps:

[0196] S31. Allocate tasks to the weighted initial regional topological matrices of each construction area and each construction task type based on the initial dynamic topological structure model to the distributed computing nodes;

[0197] S32. On each distributed computing node, extract the local weighted adjacency relationship from the weighted initial regional topological matrix, and use the normalization method to calculate the transfer probability matrix to obtain the weighted transfer probability;

[0198] S33. Define a regional optimization model for topological entropy calculation based on the weighted transfer probability, calculate the topological entropy values returned by each distributed computing node, and aggregate them to generate the global initial topological entropy. Its expression is as follows:

[0199]

[0200] Where: is the global initial topological entropy, is the topological entropy value, ψ j，k (p u，v ) = p u,v ·ξ(E′ t，j，k ) is the weighted result of the transfer probability between nodes.

[0201] The distributed iteration module 4 is used to update the initial dynamic topological structure model by using the distributed iteration method according to the time progress of the construction tasks and the real-time changes of the environmental monitoring data, and perform dynamic iterative calculation on the topological entropy values of each construction area;

[0202] Further, the distributed iteration module 4 is used to calculate the dynamic change rate according to the following steps:

[0203] S41. Based on the initial dynamic topological structure model and the topological entropy value, allocate the weighted initial regional topological matrix W j，k，t , the environmental detection data E″ t，j，k and the topological entropy value corresponding to each moment t in the initial dynamic topological structure model to the distributed computing nodes to generate a dynamic topological update task set T j，k，t ;

[0204] S42. On each distributed computing node, according to the time progress of the construction tasks and the real-time changes of the environmental monitoring data, use the initial dynamic topological structure model (Tj，k，t , W j，k，t ) and adopt a dynamic weight adjustment function to update the weighted initial regional topology matrix;

[0205] S43. Based on the updated weighted initial regional topology matrix W′ j，k，t , recalculate the topological entropy values H′ of each construction area and construction task at time t according to the topological entropy value calculation method j，k，t ;

[0206] S44. Combine the topological entropy values H′ on the time series j，k，t to calculate the dynamic change rate ΔH of the construction area and construction task j，k，t , and its expression is as follows:

[0207]

[0208] Where: Δt is the adjacent time step.

[0209] The abnormal area identification module 5 is used to calculate the topological entropy value after iterative update based on the anomaly detection algorithm, and identify the construction area where the topological entropy value has changed significantly before and after iterative update, and identify the construction progress abnormal area;

[0210] Furthermore, the abnormal area identification module 5 is used to identify the construction progress abnormal area according to the following steps:

[0211] S51. Based on the dynamic change rate ΔH j，k，t and combine the statistical characteristics of the dynamic change rate to define the anomaly detection threshold θ j，k , and its expression is as follows:

[0212] θ j，k = μ j，k + λ·σ j,k

[0213] Where: μ j，k is the mean value of the dynamic change rate under the construction area R j , construction task type Q k , σ j，k is the standard deviation of the dynamic change rate, and λ is the adjustment parameter;

[0214] S52. Compare the dynamic change rate ΔH j，k，t with the anomaly detection threshold θ j，k ; if |ΔH j，k，t | > θ j，k , it means that there is an abnormal dynamic change in the construction area R j and construction task type Q k at time t, then record the abnormal time point and abnormal change area A = {(j, k, t)||ΔHj,k,t | > θ j,k};

[0215] S53. Calculate the deviation degree of the dynamic change rate of the abnormal change area, and classify the construction progress abnormal area based on the deviation degree.

[0216] Anomaly quantification module 6 is used to determine the critical path and critical nodes of the construction tasks in combination with the critical path analysis method, and quantify the influence scope of the progress abnormal change in the construction area on the overall construction tasks.

[0217] Furthermore, the anomaly quantification module 6 is used to quantify the influence scope according to the following steps:

[0218] S61. Based on the initial dynamic topology structure model (T j，k，t , W j，k，t ) and the updated weighted initial area topology matrix W' j，k，t , construct the dynamic critical path network model CPM j，k of the construction tasks, and its expression is as follows:

[0219]

[0220] Where: (p u , p v , w' u，v ) is the dynamic weighted path from node p u to node p v in the construction tasks, and w' u，v is the path weight;

[0221] S62. Based on the dynamic critical path network model, analyze the cumulative weights of all paths through the critical path algorithm to determine the set of critical paths and nodes, and its expression is as follows:

[0222]

[0223] Where: CP j，k is the set of critical paths under the construction area R j , construction task type Q k . The path with the largest cumulative path weight is the critical path of the construction area R j , construction task type Q k ;

[0224] S63. Based on the set of critical paths CP j，k , update the topology structure of the abnormal change area (j, k, t) ∈ A, and conduct a propagation analysis on the influence on the critical path to quantify the influence scope of the progress abnormal change in the area on the overall construction tasks, and its expression is as follows:

[0225]

[0226] wherein: δCP j，k is the influence ratio of the area with abnormal construction progress on the cumulative weight of the critical path.

[0227] Embodiment 3:

[0228] Refer to Figure 3 , a non-linear dynamic monitoring device for construction progress based on topological entropy iteration, the device includes a processor 7 and a memory 8;

[0229] The memory 8 is used to store the computer program code 81 and transmit the computer program code 81 to the processor 7;

[0230] The processor 7 is used to execute the non-linear dynamic monitoring method for construction progress based on topological entropy iteration described in Embodiment 1 according to the instructions in the computer program code 81.

[0231] In this embodiment, there is also a computer-readable storage medium, and computer-executable instructions are stored in the computer-readable storage medium. When the computer-executable instructions are executed on a computer, the non-linear dynamic monitoring method for construction progress based on topological entropy iteration described in Embodiment 1 is implemented.

[0232] Generally speaking, the computer instructions for implementing the method of the present invention can be carried by any combination of one or more computer-readable storage media. A non-transitory computer-readable storage medium can include any computer-readable medium except the signal propagating temporarily itself.

[0233] The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EKROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device.

[0234] Computer program code for performing the operations of the present invention may be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and also including conventional procedural programming languages such as the "C" language or similar programming languages. In particular, the Python language suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. can be used. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., connected through the Internet using an Internet service provider).

[0235] For the above-mentioned devices and non-transitory computer-readable storage media, reference may be made to the specific description of a method for non-linear dynamic monitoring of construction progress based on topological entropy iteration and its beneficial effects, which will not be elaborated here.

[0236] Although the embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A non-linear dynamic monitoring method for construction progress based on topological entropy iteration, characterized in that, Including: S1. Collect the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site, generate a multi-source comprehensive dataset, and perform standardized preprocessing to obtain a standardized multi-source comprehensive dataset; S2. Divide the standardized multi-source comprehensive dataset into multiple sub-datasets according to the construction area, construction task, and construction time, and generate an initial dynamic topological structure model of the infrastructure construction site through a triangular mesh division algorithm based on the spatial position relationship between the sub-datasets, construction area division, and construction task; S3. Based on the distributed computing architecture, calculate the topological entropy values of each construction area in the initial dynamic topological structure model at the initial time point respectively, and aggregate them to generate the global initial topological entropy; S4. According to the time progress of the construction task and the real-time changes of the environmental monitoring data, use the distributed iterative method to update the initial dynamic topological structure model, and perform dynamic iterative calculation on the topological entropy values of each construction area; S5. Calculate the topological entropy values after iterative update based on the anomaly detection algorithm, and identify the construction areas where significant changes occur before and after the topological entropy values after iterative update, and identify the construction progress abnormal areas; S6. Combine the critical path analysis method to determine the critical path and critical nodes of the construction task, and quantify the influence range of the abnormal progress change of the construction area on the overall construction task.

2. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 1, characterized in that: The step S1 specifically includes: S11. Collect the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site; The image data includes: the construction area boundary of the infrastructure construction site, the object surface texture information, and the spatial information of the construction area; the 3D point cloud data includes the 3D spatial position and structural information of the construction area; the environmental monitoring data includes temperature, humidity, and wind speed parameters; S12. Align and fuse the image data, 3D point cloud data, and environmental monitoring data of the infrastructure construction site according to the time stamp t to generate a multi-source comprehensive dataset D containing spatial information, time information, and resource information, and its expression is as follows: D = {(I t , P t , E t ) | t ∈ [1, T]}; Where: I t is the image data at time t, P t is the 3D point cloud data at time t, E t is the environmental monitoring data at time t; S13. Denoise the image data I of the infrastructure site based on a Gaussian filter t and perform distortion correction based on the camera internal parameter matrix K and the distortion parameter vector (k1, k2, p1, p2) to obtain the corrected image data I″ t ; S14. Perform density smoothing on the three-dimensional point cloud data P of the infrastructure site t ; Adjust the density of the three-dimensional point cloud data based on the filtering algorithm of nearest neighbor search, perform three-dimensional coordinate alignment on the smoothed three-dimensional point cloud data, and perform spatial alignment through the rigid body transformation matrix T = [R|t] to obtain the aligned three-dimensional point cloud data P″ t ; S15. Perform outlier rejection on the environmental monitoring data E of the infrastructure site t and replace the data beyond the range with linear interpolation based on the set upper and lower threshold values [L min , L max to obtain the environmental monitoring data E″ after rejection t ; S16. Normalize the image data I″ t , the three-dimensional point cloud data P″ t , and the environmental monitoring data E″ t , calculate the normalization values respectively, and integrate them into a standardized multi-source comprehensive data set in a unified format. The expression is as follows: D′ = {(I″ t , P″ t , E″ t ) | t ∈ [1, T]}.

3. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 2, characterized in that: The step S2 specifically includes: S21. Divide the standardized multi-source comprehensive dataset according to the construction area, construction task, and construction time, and classify the elements that meet the specific construction area, specific construction task, and specific construction time into sub-datasets by constructing a selection function Ψ, and its expression is as follows: D′ j，k，t = Ψ(D′, R j , Q k , t) = {(I″ t，j，k , P″ t，j，k , E″ t，j，k )}; where: Ψ: D′×R j ×Q k ×t → D′ j，k，t is the partitioning operation function; I″ t，j，k is the corrected image data at time t in the construction area R j under the construction task type Q k ; P″ t，j，k is the aligned three-dimensional point cloud data under the same conditions as I″ t，j，k ; E″ t，j，k is the environmental monitoring data after outlier rejection and interpolation under the same conditions as I″ t，j，k ; S22. Based on the three-dimensional point cloud data P″ t，j，k Generate the time t, construction area R j , and the construction task type Q k The corresponding triangular mesh division result, and use the triangular mesh division algorithm to perform spatial dissection on P″ t，j，k to obtain a triangular element set, and its expression is as follows: P″ t，j，k = {p t，j，k，l = (X t,j，k，l , y t，j，k，l , z t，j，k，l ) | l ∈ [1, L j，k，t}; Where: T j，k，t is a set of triangular elements; is the three-point index forming a triangular element; U j，k,t is the total number of triangular elements generated by partitioning; is the μ-th triangular element generated at time t in construction area R j , construction task type Q k , corresponding to three point cloud point indices μ1, μ2, μ3; U j，k，t is the number of generated triangular elements; p t，j，k，l is the number of points contained in the l-th L j three-dimensional point cloud data set at time t in construction area R k , construction task type Q j，k，t ; X, y, Z are three-dimensional point cloud coordinates; S23. Based on the set T of triangular elements j，k，t Define the initial regional connection relationship and construct the initial regional connection matrix A j，k，t , and its expression is as follows: Where: a u,v is an element of the initial region connection matrix; are any two triangular elements; S24. Utilize environmental monitoring data E″ t，j，k to weight the initial regional connection matrix A j，k，t and obtain the weighted initial regional topology matrix W j，k，t , and its expression is as follows: Where: W u，v is an element of the weighted initial regional topology matrix; S25. Combine the weighted initial regional topology matrices, corresponding sub-datasets, and triangular element sets obtained at different times t, different construction areas R j , different construction task types Q k according to the time series and spatial distribution combination to obtain an initial dynamic topology structure model, and its expression is as follows: M = {(D′ j，k,t , T j，k，t , W j，k,t ) | j ∈ [1, J], k ∈ [1, K], t ∈ [1, T]}; Where: M is the initial dynamic topological structure model.

4. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 3, characterized in that: The step S3 specifically includes: S31. Allocate tasks to the weighted initial regional topological matrices of each construction area and each construction task type based on the initial dynamic topological structure model according to the distributed computing nodes; S32. On each distributed computing node, extract the local weighted adjacency relationship from the weighted initial regional topological matrix, and use the normalization method to calculate the transfer probability matrix to obtain the weighted transfer probability; S33. Based on the region optimization model for calculating topological entropy defined by weighted transfer probability, calculate the topological entropy values returned by each distributed computing node, and aggregate them to generate the global initial topological entropy. The expression is as follows: Wherein: is the global initial topological entropy, is the topological entropy value, ψ j，k (p u，v ) = p u，v ·ξ(E′ t，j，k ) is the weighted result of the transfer probability between nodes.

5. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 4, wherein: The step S4 specifically includes: S41. Based on the initial dynamic topology structure model and the topological entropy value, allocate the weighted initial regional topology matrix W corresponding to each moment t in the initial dynamic topology structure model j，k，t , the environmental detection data E″ t，j，k and the topological entropy value to the distributed computing nodes to generate a dynamic topology update task set T j，k，t ; S42. On each distributed computing node, according to the time progress of the construction task and the real-time changes of the environmental monitoring data, use the initial dynamic topology structure model (T j，k，t , W j，k，t ) and adopt the dynamic weight adjustment function to update the weighted initial regional topology matrix; S43. Based on the updated weighted initial regional topology matrix W' j，k，t , recalculate the topological entropy values H' of each construction area and construction task at time t according to the topological entropy value calculation method j，k，t ; S44. Calculate the dynamic change rate ΔH of the construction area and construction tasks in combination with the change of the topological entropy value H′ j，k，t over time series, and its expression is as follows: j，k,t ​ where: Δt is the adjacent time step.

6. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 5, wherein: The step S5 specifically includes: S51. Based on the dynamic change rate ΔH j，k，t and combining with the statistical characteristics of the dynamic change rate, define the anomaly detection threshold θ j，k , and its expression is as follows: θ j，k = μ j，k + λ·σ j，k ; Where: μ j,k is the mean of the dynamic change rate under the construction area R j , the construction task type Q k , σ is the standard deviation of the dynamic change rate, and λ is the adjustment parameter; j，k ​ S52. Compare the dynamic change rate ΔH j，k,t with the anomaly detection threshold θ j，k ; if |ΔH j，k，t | > θ j，k , it means that there is an abnormal dynamic change in the construction area R j and the construction task type Q k at time t. Then record the abnormal time point and the abnormal change area A = {(j, k, t)||ΔH j，k，t | > θ j，k}; S53. Calculate the deviation degree of the dynamic change rate of the abnormal change region, and classify the construction progress abnormal region based on the deviation degree.

7. The non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to claim 6, wherein: The step S6 specifically includes: S61. Based on the initial dynamic topology structure model (T j，k，t , W j，k，t ) and the updated weighted initial regional topology matrix W' j，k，t , construct the dynamic critical path network model CPM j，k of the construction task, and its expression is as follows: Where: (p u , p v , w′ u，v ) is the dynamic weighted path from node p u to node p v in the construction task, and w′ u，v is the path weight; S62. Based on the dynamic critical path network model, analyze the cumulative weights of all paths through the critical path algorithm, and determine the set of critical paths and nodes. The expression is as follows: Among them: CP j，k is the critical path set under the construction area R j , the construction task type Q k . The path with the largest cumulative weight of the paths is the critical path of the construction area R j , the critical path of the construction task type Qk; S63. Based on the critical path set CP j，k Update the topological structure of the abnormal change region (j, k, t) ∈ A, conduct a propagation analysis of the impact on the critical path, and quantify the impact range of the abnormal change in the progress of the region on the overall construction task. The expression is as follows: Where: δCP j，k is the influence ratio of the abnormal construction progress area on the cumulative weight of the critical path.

8. A non-linear dynamic monitoring system for construction progress based on topological entropy iteration, characterized in that, The system is applied to the method described in any one of claims 1-7. The system includes: A multi-source integrated data set construction module (1) for collecting image data, three-dimensional point cloud data and environmental monitoring data of the infrastructure site, generating a multi-source integrated data set, and performing standardized preprocessing to obtain a standardized multi-source integrated data set; An initial dynamic topological structure model construction module (2) for dividing the standardized multi-source integrated data set into multiple sub-data sets according to the construction area, construction task and construction time, and generating an initial dynamic topological structure model of the infrastructure site through a triangular mesh division algorithm based on the sub-data sets in combination with the spatial position relationship of the construction area division and construction task; A global initial topological entropy calculation module (3) for calculating the topological entropy values of each construction area in the initial dynamic topological structure model at the initial time point based on a distributed computing architecture, and aggregating them to generate the global initial topological entropy; A distributed iteration module (4) for updating the initial dynamic topological structure model by using a distributed iteration method according to the time progress of the construction task and the real-time changes of the environmental monitoring data, and performing dynamic iterative calculation on the topological entropy values of each construction area; An abnormal area identification module (5) for calculating the topological entropy values after iterative update based on an abnormal detection algorithm, and identifying the construction areas with significant changes before and after the topological entropy values after iterative update, and identifying the construction progress abnormal areas; An abnormal quantification module (6) for determining the critical path and critical nodes of the construction task by combining the critical path analysis method, and quantifying the influence range of the abnormal progress change of the construction area on the overall construction task.

9. A non-linear dynamic monitoring device for construction progress based on topological entropy iteration, wherein: The device includes a processor (7) and a memory (8); The memory (8) is used to store computer program code (81) and transmit the computer program code (81) to the processor (7); The processor (7) is configured to execute the non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to any one of claims 1-7 based on the instructions in the computer program code (81).

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed on a computer, the non-linear dynamic monitoring method for construction progress based on topological entropy iteration according to any one of claims 1-7 is implemented.

Images