Digital engineering implementation dynamic monitoring system and method based on Internet of Things

Abstract

The invention belongs to the technical field of engineering monitoring, and discloses a digital engineering implementation dynamic monitoring system and method based on the Internet of Things. The method comprises the following steps: dynamically deploying a plurality of Internet of Things sensor nodes, collecting real-time data of an engineering site, integrating the real-time data into a data stream, transmitting the data stream to a cloud server through a wireless communication network, and on the cloud server, carrying out real-time segmentation on the transmitted data stream to obtain related data groups; dividing the engineering site into a plurality of grid regions, and performing data modeling on each grid region based on the related data groups to obtain data models in the grid regions; calculating an optimal progress route of project implementation by utilizing a shortest path algorithm based on the data model in the grid region; according to the optimal progress route, dynamically adjusting an engineering implementation scheme, and feeding back the adjusted engineering implementation scheme to a site construction guidance terminal to guide construction operation; and the quality and the safety level of engineering implementation are comprehensively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering monitoring. More specifically, the present invention relates to a digital engineering implementation dynamic monitoring system and method based on the Internet of Things. Background Art

[0002] With the acceleration of the urbanization process, the construction demands for various infrastructure, public facilities, and civil buildings are increasing day by day. However, the traditional engineering implementation mode has been difficult to meet the requirements of high quality, high efficiency, and high safety in modern construction.

[0003] During the traditional engineering implementation process, the engineering site monitoring means are backward, mainly relying on manual inspections and fixed-point monitoring, with limited coverage and poor real-time performance; due to the complex and changeable engineering site environment, it is very difficult for traditional monitoring methods to detect and handle various emergencies in a timely manner, which easily leads to potential hazards in terms of quality, progress, and safety; secondly, the on-site data collection lacks systematicness and comprehensiveness; engineering construction involves many links, such as environment, safety, quality, meteorology, traffic, etc., but the existing data collection often only focuses on certain specific aspects, and it is impossible to form an overall data portrait of the engineering site, making it difficult to support scientific decision-making; furthermore, the formulation of engineering implementation plans relies too much on manual experience and lacks data support; due to the characteristics of engineering construction being one-time and non-repeatable, the actual situation of each engineering site is different, and relying too much on empirical judgment is likely to cause deviations, resulting in poor implementation effects of the plan; the engineering implementation process lacks flexibility and adaptability; once the engineering implementation plan is formulated, it is very difficult to make timely adjustments according to the actual changes on-site, which poses a great challenge to dealing with emergencies and affects the quality and efficiency of engineering implementation; finally, on-site construction guidance becomes a mere formality, lacking standardization and quantitative support, and unable to conduct refined control over complex operation processes, resulting in uneven construction quality.

[0004] In view of this, the present invention proposes a digital engineering implementation dynamic monitoring system and method based on the Internet of Things to solve the above problems. Summary of the Invention

[0005] To overcome the above defects of the prior art and to achieve the above object, the present invention provides the following technical solution: A digital engineering implementation dynamic monitoring system based on the Internet of Things, including: a dynamic deployment module for dynamically deploying a number of Internet of Things sensor nodes and using the Internet of Things sensor nodes to collect real-time data of the engineering site;

[0006] A data mid-stage processing module for integrating the collected real-time data into a data stream and transmitting the data stream to a cloud server through a wireless communication network, and on the cloud server, performing real-time segmentation on the incoming data stream to obtain relevant data groups;

[0007] A model establishment module, which is used to divide the engineering site into several grid areas, and perform data modeling on each grid area based on relevant data groups to obtain a data model within the grid area;

[0008] A scheme optimization module, based on the data model within the grid area, uses the shortest path algorithm to calculate the optimal progress route for project implementation; according to the optimal progress route, dynamically adjusts the project implementation plan, and feeds back the adjusted project implementation plan to the on-site construction guidance terminal; each module is connected by wired and / or wireless means.

[0009] Furthermore, the deployment method of the Internet of Things sensor nodes includes:

[0010] Obtain the floor plan of the engineering site and determine the scope of the monitoring area; divide the monitoring area into several triangular sub-areas to form an initial grid; the initial grid contains several triangular sub-areas; according to the monitoring requirements, determine the type and quantity of sensors to be deployed in each triangular sub-area; use Heron's formula or the vector cross product method to calculate the area of each triangular sub-area; calculate the inradius r and circumradius R of the triangular sub-area; obtain the minimum angle θmin and maximum angle θmax of the triangle;

[0011] Based on the area S of the triangular sub-area, calculate the shape factor SF of the triangular sub-area;

[0012] where a, b, and c are the side lengths of the three sides of the triangular sub-area;

[0013] Calculate the complexity index where w1, w2, and w3 are complexity weight coefficients;

[0014] According to the complexity index CI and the area S of the triangular sub-area, calculate the deployment priority Arrange all triangular sub-areas in descending order of the value of the deployment priority Pr;

[0015] Initialize a virtual sensor at the centroid position of each triangular sub-area; obtain the initial coverage range of each virtual sensor;

[0016] Traverse the sorted triangular sub-areas, check the coverage of the triangular sub-areas by the initial coverage range of the virtual sensors. If it is completely covered, skip the corresponding triangular sub-area; if it is partially covered or not covered, find candidate positions within the triangular sub-area; for each candidate position, calculate the newly added initial coverage range; select the position with the largest newly added initial coverage range as the optimal position to deploy the Internet of Things sensor nodes within the corresponding triangular sub-area.

[0017] Further, the method for obtaining the initial grid includes:

[0018] Obtain the boundary contour of the monitoring area. According to the preset grid density parameter, discretize the boundary contour into a series of boundary points, and connect these boundary points to form initial boundary edges;

[0019] Define a doubly linked list, add all boundary edges to the doubly linked list, and calculate an ideal triangle height h for each boundary edge in the doubly linked list;

[0020] where f is a preset global scale factor, k is the local curvature, d is the preset grid density, I is the importance index; α, β, and γ are weight coefficients, and ε is the importance influence coefficient;

[0021] Select a boundary edge from the doubly linked list as the current active edge. Generate a triangle for the current active edge based on the calculated ideal triangle height h, that is, calculate the position of the third vertex of the triangle. Set a threshold distance, and search whether there are existing grid points within the threshold distance around this vertex position. If so, use the nearest grid point; if not, create a new grid point; Check whether the triangle intersects with the existing grid elements. If it intersects, adjust the shape of the triangle. If a triangle cannot be generated, merge adjacent boundary edges; If a triangle is generated, add this triangle to the grid and remove the current active edge; Then an initial grid is obtained.

[0022] Further, the real-time data at the construction site includes environmental data, safety data, quality data, meteorological data, and traffic data.

[0023] Further, the method for integrating the collected real-time data into a data stream includes:

[0024] Perform preliminary processing on the real-time data, including denoising, outlier detection, and data format standardization; Set a time window with a fixed size and determine the step size for the time window to slide; Initially classify the real-time data according to the sensor type and location of the IoT sensor nodes. Within each sliding time window, aggregate the real-time data of the same sensor type and location together to obtain a data group; Assign a unified timestamp to each data group; Within each time window, perform an aggregation operation on the data group; Organize the data groups after the aggregation operation into a structured format, JSON or a custom data structure; Each data structure contains a timestamp, a sensor identifier, location information, and the data value after the aggregation operation; Concatenate the continuous data structures in chronological order to form a continuous data stream.

[0025] Further, the method for real-time segmentation of the incoming data stream includes:

[0026] Set segmentation parameters, where the segmentation parameters include a target value μ0, an allowable offset δ, and a decision interval; initialize the upward cumulative sum S + and the downward cumulative sum S - to 0.

[0027] Define a sliding window and use an exponentially weighted moving average to update the target value μ0. Calculate the standard deviation using the data in the sliding window, and set the allowable offset δ as a multiple of the standard deviation; set an update interval U, and perform an update of the target value μ0 and the allowable offset δ every U data points are processed;

[0028] Traverse the data stream. For the i-th newly arrived data point xi, calculate the upper deviation di = xi - (μ0 + δ); update the upward cumulative sum based on the upper deviation di. The update formula for the upward cumulative sum is:

[0029] S + _new = max(0, S + + di); where S + _new is the updated upward cumulative sum;

[0030] For the i-th newly arrived data point xi, calculate the lower deviation fi = (μ0 - δ) - xi; update the downward cumulative sum based on the lower deviation fi. The update formula for the downward cumulative sum is:

[0031] S - _new = max(0, S - + fi) where S - _new is the updated downward cumulative sum;

[0032] If S + _new is greater than the decision interval or S - _new is greater than the decision interval, then a change point is detected. Record the timestamp of the change point and the corresponding data value; reset the upward cumulative sum S + and the downward cumulative sum S - to 0, and perform detection on the (i + 1)-th newly arrived data point; based on the detected change point, segment the data stream into several segments; assign a unique identifier to each segment, and record the start and end timestamps;

[0033] Extract the time series features of each segment, and calculate the correlation coefficient between adjacent segments using the sliding window algorithm based on the time series features; preset a correlation threshold, and combine adjacent segments with a correlation coefficient greater than the correlation threshold into a correlated data group.

[0034] Further, the method of dividing the engineering site into several grid regions includes:

[0035] Obtain the geographical information data of the engineering site; project the geographical information data onto a unified coordinate system, and set the side length d0 of the initial grid area; select a corner point of the engineering site boundary as the starting point, and starting from the starting point, generate a regular square grid array according to the set side length d0. Mark the area that cannot be completely covered by the regular square grid array as a residual polygon;

[0036] For each residual polygon, calculate its area A. If A > d0 2 , then halve d0 and repeat generating a regular square grid array inside the residual polygon; if A ≤ d0 2 , then regard the residual polygon as an independent grid area and do not subdivide it anymore;

[0037] Scan all the square grids, find the square grids with side length less than d0, and record them as small grids; merge the adjacent small grids. The merging condition is that the side length after merging does not exceed d0; thus, obtain several grid areas.

[0038] Furthermore, the method of data modeling for each grid area includes:

[0039] Obtain the geographical location information of the data points in each relevant data group, traverse all grid areas, and for each grid area, check the coordinates of its boundary, and associate the data points falling within the boundary of the grid area and their affiliated relevant data groups with the corresponding grid area;

[0040] For each grid area, extract all the associated relevant data groups within it and construct them into a time series; perform differencing on the time series to obtain a differenced series until the differenced series satisfies stationarity;

[0041] Draw the autocorrelation graph of the differenced series. If the autocorrelation graph shows a trailing phenomenon and the autocorrelation coefficient values gradually decay to 0 only after the p-th order, then select the autoregressive model as the initial modeling model; if the autocorrelation graph truncates to 0 after the q-th order, then select the moving average model as the initial modeling model. If the autocorrelation graph both shows a trailing phenomenon and truncates to 0 after the q-th order, then select the autoregressive moving average model as the initial modeling model; p and q are both the orders of the initial modeling model; use the least squares method or the maximum likelihood estimation method to estimate the parameters of the initial modeling model and obtain the residual series; check whether the residual series is non-autocorrelated, has a mean of 0, and a constant variance; if the residual series is non-autocorrelated, has a mean of 0, and a constant variance, then record this residual series as the conditional series;

[0042] Based on the conditional series, construct a data model for the corresponding grid area. The expression of the data model is:

[0043] σ(t) 2 = a0 + ∑(s(u) × |∈(t - u)|x1 ) + ∑(e(v) × σ(t - v) 2 ) + ρ × D_t; wherein

[0044] wherein, σ(t) 2 is the conditional variance at time t, a0 is a constant term greater than 0; s(u) is the autoregressive term coefficient; x1 is the exponential coefficient; e(v) is the generalized autoregressive term coefficient; σ(t - v) 2 is the conditional variance at the past time t - v; ∈(t - u) is the value of the residual term representing the time series at time t - u; u is the time lag order of the autoregressive term, v is the time lag order of the generalized autoregressive term; D_t is a 0 - 1 dummy variable; ρ is the coefficient of the dummy variable.

[0045] Furthermore, the method for obtaining the optimal progress route includes:

[0046] Regarding the grid regions divided in the engineering site as nodes in an undirected graph, for adjacent grid regions, connect an undirected edge between them, and set the weight of each undirected edge as the difference in the conditional variances of the two adjacent grid regions at the future time point T;

[0047] Select a starting point and an ending point in the grid regions. If there are multiple starting points / ending points, connect them to the adjacent grid regions with undirected edges having a weight of 0;

[0048] Run the shortest path algorithm on the undirected graph, with the starting point as the source point, find the shortest path from the starting point to the ending point, and the sequence formed by the nodes on the shortest path is the preliminary optimal progress route for the project implementation; perform smoothing processing on the preliminary optimal progress route to obtain the optimal progress route.

[0049] The dynamic monitoring method for digital project implementation based on the Internet of Things is implemented based on the described dynamic monitoring system for digital project implementation based on the Internet of Things, and includes: Step 1, dynamically deploy a number of Internet of Things sensor nodes, and use the Internet of Things sensor nodes to collect real - time data of the engineering site;

[0050] Step 2, integrate the collected real - time data into a data stream, and transmit the data stream to the cloud server through a wireless communication network. On the cloud server, perform real - time segmentation on the incoming data stream to obtain relevant data groups;

[0051] Step 3, divide the engineering site into several grid regions, and perform data modeling on each grid region based on the relevant data groups to obtain the data model within the grid region;

[0052] Step 4: Based on the data model within the grid area, use the shortest path algorithm to calculate the optimal progress route for project implementation; according to the optimal progress route, dynamically adjust the project implementation plan, and feedback the adjusted project implementation plan to the on-site construction guidance terminal.

[0053] Technical effects and advantages of the digital project implementation dynamic monitoring system and method based on the Internet of Things of the present invention:

[0054] The present invention brings all-round intelligent monitoring and refined management to project implementation, greatly improving the quality, efficiency and safety of project implementation. First, it realizes full coverage and real-time dynamic data collection at the project site, eliminates monitoring blind spots, and provides accurate and timely data support for subsequent decision-making; second, through data-driven modeling and analysis, it deeply explores the internal laws of the project site, effectively avoids human experience deviations in the traditional way, and improves the scientificity and forward-looking of decision-making; third, based on the differences in data models, it automatically calculates the optimal progress route for project implementation, minimizing the cost and risk of project conversion to the greatest extent and ensuring the efficient and orderly progress of project implementation; in addition, the project implementation plan can be dynamically optimized and adjusted according to the on-site real-time data, improving the adaptability and flexibility of project implementation and quickly responding to various emergencies; finally, the optimized project implementation plan is decomposed into specific operation guides for each construction area, providing refined and standardized operation processes for on-site construction, thus comprehensively improving the quality and safety level of project implementation. Description of the Drawings

[0055] Figure 1 It is a schematic diagram of the digital project implementation dynamic monitoring system based on the Internet of Things of the present invention;

[0056] Figure 2 It is a schematic diagram of the digital project implementation dynamic monitoring method based on the Internet of Things of the present invention. Detailed Embodiments

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0058] Embodiment 1

[0059] Please refer to Figure 1 As shown, the digital project implementation dynamic monitoring system based on the Internet of Things in this embodiment includes:

[0060] A dynamic deployment module for dynamically deploying a number of Internet of Things (IoT) sensor nodes to collect real-time data at the engineering site using the IoT sensor nodes;

[0061] A data mid-processing module for integrating the collected real-time data into a data stream and transmitting it to a cloud server via a wireless communication network. At the cloud server, the incoming data stream is segmented in real time to obtain relevant data groups;

[0062] A model building module for dividing the engineering site into several grid areas and performing data modeling on each grid area based on the relevant data groups to obtain a data model within the grid area;

[0063] A scheme optimization module, based on the data model within the grid area, uses the shortest path algorithm to calculate the optimal progress route for the project implementation; according to the optimal progress route, dynamically adjusts the project implementation plan and feeds back the adjusted project implementation plan to the on-site construction guidance terminal to guide the construction operation; each module is connected by wired and / or wireless means to achieve data transmission between modules.

[0064] The deployment methods of the IoT sensor nodes include:

[0065] Obtain a floor plan of the engineering site and determine the scope of the monitoring area; divide the monitoring area into several triangular sub-areas to form an initial grid; specifically, obtain the boundary contour of the monitoring area, discretize the boundary contour into a series of boundary points according to a preset grid density parameter, and connect these boundary points to form an initial boundary edge.

[0066] Define a doubly linked list, add all the boundary edges to the doubly linked list, and calculate an ideal triangle height h for each boundary edge in the doubly linked list;

[0067] Where f is a preset global scale factor for overall control of the grid size, k is the local curvature, reflecting the degree of curvature of the boundary contour at the location of the boundary edge; d is the preset grid density (which can be understood as the desired average side length), I is the importance index, an integer greater than 1 determined by relevant personnel based on factors such as project criticality and safety requirements. The important areas (with a larger I) will get a denser grid division; α, β, and γ are weight coefficients for balancing the effects of curvature, density, and minimum size; ε is the importance influence coefficient for balancing the influence of the importance index.

[0068] Select a boundary edge from the doubly linked list as the current active edge, and generate a triangle for the current active edge based on the calculated ideal triangle height h, that is, calculate the position of the third vertex of the triangle. Set a threshold distance (such as 10% of the triangle height h), and search whether there are existing grid points within the threshold distance around this vertex position. If there are, use the nearest grid point; if not, create a new grid point (create a new grid point based on the calculated ideal triangle height); check whether the triangle intersects with the existing grid elements. If it intersects, adjust the shape of the triangle. If a triangle cannot be generated, merge adjacent boundary edges; if a triangle is generated, add this triangle to the grid and remove the current active edge; thus, an initial grid is obtained.

[0069] It should be noted that grid points are the vertices of triangles, including existing points (such as boundary points and previously generated internal points) and newly generated points; the position of the third vertex of the ideal triangle is based on the current active edge and the calculated ideal triangle height to determine the ideal position of a new point, in the normal direction of the midpoint of the current active edge, and the distance is the calculated ideal triangle height; checking whether the triangle intersects with the existing grid elements can be understood as whether the newly generated triangle overlaps with the existing triangles, including the cases of edge intersection and points falling inside other triangles.

[0070] The initial grid contains several triangles, denoted as triangle sub-regions; according to the monitoring requirements, determine the types and quantities of sensors to be deployed within each triangle sub-region; the sensor types include temperature and humidity sensors, noise sensors, dust sensors, etc.

[0071] Use Heron's formula or the vector cross product method to calculate the area of each triangle sub-region; calculate the inradius r and circumradius R of the triangle sub-region; obtain the minimum angle θmin and maximum angle θmax of the triangle; based on the area S of the triangle sub-region, calculate the shape factor SF of the triangle sub-region;

[0072] where a, b, and c are the side lengths of the three sides of the triangle sub-region;

[0073] Calculate the complexity index where w1, w2, and w3 are complexity weight coefficients; the larger the value of the complexity index, the more irregular the triangle and the higher the complexity;

[0074] According to the complexity index CI and the area S of the triangle sub-region, calculate the deployment priority Arrange all triangle sub-regions in descending order of the value of the deployment priority Pr.

[0075] Initialize a virtual sensor at the centroid position of each triangular sub-region; obtain the initial coverage range of each virtual sensor. Specifically, if it is an omnidirectional sensor, draw a circle with the sensor position as the center and the maximum sensing distance R_max of the sensor as the radius; if it is a directional sensor, draw a sector with the sensor position as the vertex, R_max as the radius, and the sensing angle range θ as the angle.

[0076] Traverse the sorted triangular sub-regions to check the coverage of the triangular sub-regions by the initial coverage range of the virtual sensors. If it is completely covered, skip the corresponding triangular sub-region; if it is partially covered or not covered, find candidate positions within the triangular sub-region (considering vertices and midpoints of sides of the triangle, etc.); for each candidate position, calculate the newly added initial coverage range; select the position with the largest newly added initial coverage range as the optimal position for deploying the Internet of Things sensor node within the corresponding triangular sub-region; generate the final sensor deployment plan, including the specific coordinate positions and types of each sensor; during the engineering implementation process, dynamically adjust the sensor deployment according to changes in the on-site situation, such as adding, removing, or repositioning sensor nodes.

[0077] The real-time data at the engineering site includes environmental data, safety data, quality data, meteorological data, and traffic data.

[0078] Environmental data includes temperature, humidity, noise level, dust concentration, air quality (such as CO2 concentration, VOC, etc.), and light intensity; safety data includes harmful gas concentration (such as CO, H2S, etc.), vibration intensity, tilt angle (for monitoring structural stability), and crack width (for monitoring structural deformation); quality data includes concrete strength, steel stress, structural deformation amount, and settlement amount; meteorological data includes wind speed, wind direction, and rainfall; traffic data includes vehicle entry and exit frequency and the congestion rate per unit time of the construction road.

[0079] These data types cover multiple aspects during the engineering construction process, helping to monitor the project progress, quality, safety, and environmental impact in real time; the specific data types and collection frequencies are adjusted according to the specific requirements and importance of the project.

[0080] Continuously collect real-time data from various IoT sensor nodes, and perform preliminary processing on the original real-time data, including denoising, outlier detection, and data format standardization; Set a time window of a fixed size, such as 5 minutes, 10 minutes, or 15 minutes, and determine the step size for the time window to slide, which can be half of the window size or other appropriate values; Initially classify the real-time data according to the sensor type and location of the IoT sensor nodes. Within each sliding time window, aggregate the real-time data of the same sensor type and location together to obtain a data group; Assign a unified timestamp to each data group, usually using the start or midpoint time of the time window; Within each time window, perform aggregation operations on the data group, such as calculating the average value, maximum value, minimum value, or other statistics; Organize the data group after the aggregation operation into a structured format, such as JSON or a custom data structure, and each data structure contains a timestamp, sensor identifier, location information, and the data value after the aggregation operation.

[0081] Concatenate the continuous data structures in chronological order to form a continuous data stream; Ensure that each data group in the data stream has a unique identifier for subsequent processing and tracking; Use an appropriate communication protocol (such as MQTT, CoAP, or a custom protocol) to transmit the data stream to the cloud server; Implement an encryption and security mechanism for data transmission to protect sensitive information.

[0082] Effectively integrate the scattered and heterogeneous sensor data into a continuous and structured data stream, providing a reliable data basis for subsequent real-time analysis and decision-making; It can not only ensure the timeliness of data but also improve the efficiency of data processing.

[0083] The ways to perform real-time segmentation on the incoming data stream include:

[0084] Set segmentation parameters, where the segmentation parameters include the target value μ0 (usually the expected mean of the data), the allowable offset δ (adjusted according to the engineering implementation scenario), and the decision interval (controlling the sensitivity); Initialize the cumulative sum S + for the upward and the cumulative sum S - for the downward to be 0.

[0085] Define a sliding window and use the exponentially weighted moving average (EWMA) to update the target value μ0. Calculate the standard deviation σ using the data in the sliding window, and set the allowable offset δ as a multiple of the standard deviation, such as δ = k1×σ, where k1 is an adjustable parameter (such as k = 2 or 3); Set an update interval U (such as updating once every 50 data points processed). After processing U data points, perform an update of the target value μ0 and the allowable offset δ.

[0086] Traverse the data stream. For the \(i\)-th newly arrived data point \(x_i\), calculate the upper deviation \(d_i = x_i - (\mu_0+\delta)\); update the upper cumulative sum based on the upper deviation \(d_i\). The update formula for the upper cumulative sum is:

[0087] S + _new = max(0, S + +d_i); where S + _new is the updated upper cumulative sum;

[0088] For the \(i\)-th newly arrived data point \(x_i\), calculate the lower deviation \(f_i = (\mu_0 - \delta)-x_i\); update the lower cumulative sum based on the lower deviation \(f_i\). The update formula for the lower cumulative sum is:

[0089] S - _new = max(0, S - +f_i) where S - _new is the updated lower cumulative sum;

[0090] If S + _new is greater than the decision interval or S - _new is greater than the decision interval, then a change point is detected. Record the timestamp of the change point and the corresponding data value; reset the upper cumulative sum S + and the lower cumulative sum S - to 0, and perform detection on the \((i + 1)\)-th newly arrived data point; based on the detected change points, divide the data stream into several segments; assign a unique identifier to each segment, and record the start and end timestamps.

[0091] Extract the time series features of each segment (such as autocorrelation coefficient, spectral features, etc.), and calculate the correlation coefficient between adjacent segments using the sliding window algorithm based on the time series features; preset a correlation threshold, and combine adjacent segments with a correlation coefficient greater than the correlation threshold into a correlated data group; assign a unique identifier to each correlated data group, and record the IDs and time ranges of the segments within the data segment group.

[0092] Identifying correlated data groups can not only quickly capture important change points in the data, but also adapt to a dynamically changing environment, providing reliable data support for real-time monitoring and decision-making in the engineering field.

[0093] The ways to divide the engineering site into several grid regions include:

[0094] Obtain the geographical information data of the engineering site (obtained from remote sensing images, geographic information system (GIS) data, etc.); project the geographical information data onto a unified coordinate system, and set the side length \(d_0\) of the initial grid region. \(d_0\) is usually set to 10 - 50 meters and can be adjusted according to specific circumstances.

[0095] Select a corner point of the project site boundary as the starting point. Starting from the starting point, generate a regular square grid array according to the set side length d0. Mark the area that cannot be completely covered by the regular square grid array as the residual polygon;

[0096] For each residual polygon, calculate its area A. If A > d0 2 (initial grid area), then halve d0 and repeat generating a regular square grid array inside the residual polygon; if A ≤ d0 2 , then regard the residual polygon as an independent grid area and do not subdivide it anymore;

[0097] Scan all square grids, find the square grids with side lengths less than d0, denoted as small grids; merge adjacent small grids; the merging condition is that the side length after merging does not exceed d0; then obtain several grid areas; number all grid areas according to a certain rule (such as row-first or column-first) and assign a unique ID; record the boundary coordinates (lower left corner and upper right corner) of each grid area for subsequent spatial association.

[0098] The ways to perform data modeling for each grid area include:

[0099] Obtain the geographical location information (latitude and longitude coordinates or projected coordinates) of the data points in each relevant data group. Traverse all grid areas. For each grid area, check the coordinates of its boundary and associate the data points falling within the boundary of the grid area and their corresponding relevant data groups with the corresponding grid area;

[0100] For each grid area, extract all associated relevant data groups within it and construct them into a time series; perform differencing on the time series (perform first-order differencing or seasonal differencing) to obtain a differenced series until the differenced series satisfies stationarity; the test for stationarity can use the unit root test (ADF test) or other statistical test methods.

[0101] Plot the autocorrelation function (ACF) of the difference sequence. If the autocorrelation plot shows a trailing phenomenon and the autocorrelation coefficient values gradually decay to 0 only after the p-th order, then select the autoregressive model (AR) as the initial modeling model; if the autocorrelation plot truncates to 0 after the q-th order, then select the moving average model (MA) as the initial modeling model. If the autocorrelation plot shows both a trailing phenomenon and truncates to 0 after the q-th order, then select the autoregressive moving average model (ARMA) as the initial modeling model; both p and q are the orders of the initial modeling model. Use methods such as the least squares method or maximum likelihood estimation to estimate the parameters of the initial modeling model and obtain the residual sequence. Check whether the residual sequence is autocorrelation-free, has a mean of 0, and a constant variance (Ljung-Box test, Jarque-Bera normality test, etc. can be used). If the residual sequence is autocorrelation-free, has a mean of 0, and a constant variance, then record this residual sequence as the conditional sequence.

[0102] Based on the conditional sequence, construct a data model for the corresponding grid region. The expression of the data model is:

[0103] σ(t) 2 = a0 + ∑(s(u) × |∈(t - u)| x1 ) + ∑(e(v) × σ(t - v) 2 ) + ρ × D_t; where

[0104] σ(t) 2 is the conditional variance at time t, a0 is a constant term greater than 0 to ensure that the conditional variance is positive. s(u) is the autoregressive term coefficient, with a value greater than 0. It reflects the influence degree of the residual term ∈(t - u) on the current conditional variance. Specifically, s(u) captures the influence of the new information of the residual term on future volatility. x1 is the exponential coefficient, usually taking 1 or 2. e(v) is the generalized autoregressive term coefficient, with a value greater than 0, reflecting the influence degree of the past conditional variance on the current conditional variance. Specifically, it captures the characteristics of the persistence and clustering of volatility. σ(t - v) 2 is the conditional variance at the past time t - v. ∈(t - u) is the value of the residual term (residual / error term) of the time series at time t - u. u is the time lag order of the autoregressive term, v is the time lag order of the generalized autoregressive term. D_t is a 0-1 dummy variable used to capture the influence of specific events on volatility, marking whether a specific event (such as a holiday, major accident, etc.) occurs. ρ is the coefficient of the dummy variable, reflecting the influence degree of the event on volatility.

[0105] Regard the grid areas divided in the construction site as nodes in an undirected graph. For adjacent grid areas, connect an undirected edge between them, and set the weight of each undirected edge as the difference in the conditional variances of the two adjacent grid areas at the future time point T; According to the actual situation of the project implementation, select the starting point and the ending point in the grid area. If there are multiple starting points / ending points, connect them to the adjacent grid areas with undirected edges with a weight of 0.

[0106] Run the shortest path algorithm on the undirected graph, such as Dijkstra's algorithm, with the starting point as the source point to find the shortest path from the starting point to the ending point. The sequence of nodes on the shortest path constitutes the preliminary optimal progress route for the project implementation; Since the grid area is a regular square, there may be some unnatural corners in the shortest path, and smooth the shortest path to make it more natural and smooth; The smoothing methods include cubic spline interpolation, wavelet transform, etc.

[0107] Use the shortest path algorithm to calculate the optimal progress route for the project implementation based on the data model differences within the grid area; This route can minimize the environmental differences between adjacent work areas, thereby reducing the costs and risks of project conversion and improving the construction efficiency and quality.

[0108] Parse the optimal progress route into a series of grid area sequences according to the order of the grid areas. According to the grid area sequence, plan the operation content, time arrangement, required resources, etc. of each construction area of the project; Analyze the overall project implementation plan to identify possible conflicts, bottlenecks or risk points; For the areas with problems, adjust the operation arrangement order, resource allocation, etc. according to the data model to optimize the overall implementation plan; Decompose the optimized project implementation plan into specific operation guidance plans for each construction area.

[0109] This embodiment brings all-round intelligent monitoring and refined management to the project implementation, greatly improving the quality, efficiency and safety of the project implementation. First, it realizes full coverage and real-time dynamic data collection of the construction site, eliminates monitoring blind spots, and provides accurate and timely data support for subsequent decision-making; Secondly, through data-driven modeling and analysis, it deeply explores the internal laws of the construction site, effectively avoids human experience biases in the traditional way, and improves the scientificity and forward-looking of decision-making; Moreover, based on the data model differences, it automatically calculates the optimal progress route for the project implementation, minimizing the costs and risks of project conversion to the greatest extent and ensuring the efficient and orderly progress of the project implementation; In addition, the project implementation plan can be dynamically optimized and adjusted according to the on-site real-time data, improving the adaptability and flexibility of the project implementation and quickly responding to various emergencies; Finally, decompose the optimized project implementation plan into specific operation guides for each construction area, providing refined and standardized operation processes for on-site construction, thus comprehensively improving the quality and safety level of the project implementation.

[0110] Example 2

[0111] Please refer to Figure 2 as shown. For the parts not described in detail in this embodiment, refer to the description in Embodiment 1. A method for dynamically monitoring the implementation of a digital project based on the Internet of Things is provided, including:

[0112] Step 1: Dynamically deploy a number of Internet of Things sensor nodes, and use the Internet of Things sensor nodes to collect real-time data at the project site;

[0113] Step 2: Integrate the collected real-time data into a data stream, and transmit the data stream to the cloud server through a wireless communication network. On the cloud server, the incoming data stream is segmented in real time to obtain relevant data groups;

[0114] Step 3: Divide the project site into several grid areas, and perform data modeling on each grid area based on the relevant data groups to obtain a data model within the grid area;

[0115] Step 4: Based on the data model within the grid area, use the shortest path algorithm to calculate the optimal progress route of the project implementation; according to the optimal progress route, dynamically adjust the project implementation plan, and feedback the adjusted project implementation plan to the on-site construction guidance terminal to guide the construction operation.

[0116] Example 3

[0117] This embodiment publicly provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the above-provided method for dynamically monitoring the implementation of a digital project based on the Internet of Things.

[0118] Since the electronic device introduced in this embodiment is the electronic device used to implement the method for dynamically monitoring the implementation of a digital project based on the Internet of Things in the embodiments of the present application, based on the method for dynamically monitoring the implementation of a digital project based on the Internet of Things introduced in the embodiments of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in the embodiments of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device used to implement the method for dynamically monitoring the implementation of a digital project based on the Internet of Things in the embodiments of the present application, it falls within the scope of protection of the present application.

[0119] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by software simulation of a large amount of collected data to obtain a formula closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0120] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions within the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as within the protection scope of the present invention.

Claims

1. A dynamic monitoring system for digital engineering implementation based on the Internet of Things, characterized in that, Including: A dynamic deployment module for dynamically deploying a number of Internet of Things (IoT) sensor nodes to collect real-time data of the engineering site using the IoT sensor nodes; A data mid-processing module for integrating the collected real-time data into a data stream and transmitting the data stream to a cloud server via a wireless communication network. On the cloud server, the incoming data stream is segmented in real time to obtain relevant data groups; A model establishment module for dividing the engineering site into a number of grid areas and performing data modeling on each grid area based on the relevant data groups to obtain a data model within the grid area; A solution optimization module, based on the data model within the grid area, calculates the optimal progress route of the project implementation using the shortest path algorithm; according to the optimal progress route, dynamically adjusts the project implementation plan and feeds back the adjusted project implementation plan to the on-site construction guidance terminal; each module is connected by wired and / or wireless means.

2. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 1, characterized in that, The deployment method of the IoT sensor nodes includes: Obtain a floor plan of the engineering site and determine the scope of the monitoring area; divide the monitoring area into a number of triangular sub-areas to form an initial grid; the initial grid contains a number of triangular sub-areas; according to the monitoring requirements, determine the type and quantity of sensors to be deployed within each triangular sub-area; use Heron's formula or the vector cross product method to calculate the area of each triangular sub-area; calculate the inradius r and circumradius R of the triangular sub-area; obtain the minimum angle θmin and maximum angle θmax of the triangle; Based on the area S of the triangular sub-area, calculate the shape factor SF of the triangular sub-area; where a, b, and c are the side lengths of the three sides of the triangular sub-region; Computational complexity index where w1, w2, and w3 are complexity weight coefficients; Calculate the deployment priority based on the complexity index CI and the area S of the triangular sub-region Arrange all the triangular sub-regions in descending order of the value of the deployment priority Pr; Initialize a virtual sensor at the centroid position of each triangular sub-area; obtain the initial coverage range of each virtual sensor; Traverse the sorted triangular sub-areas to check the coverage of the triangular sub-areas by the initial coverage range of the virtual sensors. If it is fully covered, skip the corresponding triangular sub-area; if it is partially covered or not covered, find candidate positions within the triangular sub-area; for each candidate position, calculate the newly added initial coverage range; select the position with the largest newly added initial coverage range as the optimal position for deploying IoT sensor nodes within the corresponding triangular sub-area.

3. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 2, characterized in that, The method for obtaining the initial grid includes: Obtain the boundary contour of the monitoring area, discretize the boundary contour into a series of boundary points according to the preset grid density parameter, and connect these boundary points to form an initial boundary edge; Define a doubly linked list, add all boundary edges to the doubly linked list, and calculate an ideal triangle height h for each boundary edge in the doubly linked list; Among them, f is a preset global scale factor, k is the local curvature, d is the preset grid density, and I is the importance index; α, β, and γ are weight coefficients, and ε is the importance influence coefficient; Select a boundary edge from the doubly linked list as the current active edge, generate a triangle for the current active edge based on the calculated ideal triangle height h, that is, calculate the position of the third vertex of the triangle, set a threshold distance, and search whether there are existing grid points within the threshold distance around this vertex position. If so, use the nearest grid point; if not, create a new grid point; check whether the triangle intersects with the existing grid elements. If it intersects, adjust the shape of the triangle. If a triangle cannot be generated, merge adjacent boundary edges; if a triangle is generated, add this triangle to the grid and remove the current active edge; thus, an initial grid is obtained.

4. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 1, characterized in that, The real-time data of the construction site includes environmental data, safety data, quality data, meteorological data, and traffic data.

5. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 3, characterized in that, The methods for integrating the collected real-time data into a data stream include: Perform preliminary processing on the real-time data, including denoising, outlier detection, and data format standardization; set a time window of a fixed size and determine the step size for the time window to slide; perform preliminary classification of the real-time data according to the sensor type and location of the IoT sensor nodes. Within each sliding time window, aggregate the real-time data of the same sensor type and location to obtain a data group; assign a unified timestamp to each data group; perform an aggregation operation on the data groups within each time window; organize the data groups after the aggregation operation into a structured format, JSON or a custom data structure; each data structure includes a timestamp, a sensor identifier, location information, and the data value after the aggregation operation; concatenate the continuous data structures in chronological order to form a continuous data stream.

6. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 5, characterized in that The methods for performing real-time segmentation on the incoming data stream include: Set the segmentation parameters, which include the target value μ0, the allowed offset δ, and the decision interval; initialize the uplink accumulation and S + and the downward cumulative sum S - is 0; Define a sliding window and use an exponentially weighted moving average to update the target value μ0. Calculate the standard deviation using the data in the sliding window, and set the allowable offset δ as a multiple of the standard deviation; set an update interval U, and perform an update of the target value μ0 and the allowable offset δ every U data points are processed. Traverse the data stream. For the i-th newly arrived data point xi, calculate the upper deviation di = xi - (μ0 + δ); update the upward cumulative sum based on the upper deviation di. The update formula for the upward cumulative sum is: S + _new = max(0, S + + d_i); where, S + _new is the updated cumulative sum in the upward direction; For the i-th newly arrived data point xi, calculate the lower deviation fi = (μ0 - δ) - xi; update the downward cumulative sum based on the lower deviation fi; the update formula for the downward cumulative sum is: S - _new = max(0, S - + f_i), where S - _new is the updated cumulative sum in the downward direction; If S + _new is greater than the decision interval or S - _new is greater than the decision interval, a change point is detected, and the timestamp of the change point and the corresponding data value are recorded; the upward cumulative sum S + and the downward cumulative sum S - are reset to 0, and the (i + 1)-th newly arrived data point is detected; based on the detected change point, the data stream is segmented into several segments; a unique identifier is assigned to each segment, and the start and end timestamps are recorded; Extract the time series features of each segment, and calculate the correlation coefficient between adjacent segments using the sliding window algorithm based on the time series features; preset a correlation threshold, and combine adjacent segments with a correlation coefficient greater than the correlation threshold into a correlated data group.

7. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 6, characterized in that, The methods for dividing the construction site into several grid regions include: Obtain the geographical information data of the engineering site; project the geographical information data onto a unified coordinate system, and set the side length d0 of the initial grid area; select a corner point of the engineering site boundary as the starting point, and starting from the starting point, generate a regular square grid array according to the set side length d0. Mark the area that cannot be completely covered by the regular square grid array as a residual polygon; For each remaining polygon, calculate its area A. If A > d0 2 , then halve d0 and repeat generating a regular square grid array inside the remaining polygon; if A ≤ d0 2 , then regard the remaining polygon as an independent grid area and do not further subdivide it; Scan all the square grids, find the square grids with side lengths less than d0, and record them as small grids; merge adjacent small grids; the merging condition is that the side length after merging does not exceed d0; thus, obtain several grid areas.

8. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 7, wherein, The method of data modeling for each grid area includes: Obtain the geographical location information of the data points in each relevant data group, traverse all grid areas, for each grid area, check the coordinates of its boundary, and associate the data points falling within the boundary of the grid area and their affiliated relevant data groups with the corresponding grid area; For each grid area, extract all the associated relevant data groups within it and construct them into a time series; perform differencing on the time series to obtain a differenced series until the differenced series satisfies stationarity; Plot the autocorrelation graph of the differenced series. If the autocorrelation graph shows a trailing phenomenon and the autocorrelation coefficient values gradually decay to 0 only after the p-th order, then select the autoregressive model as the initial modeling model; if the autocorrelation graph truncates to 0 after the q-th order, then select the moving average model as the initial modeling model. If the autocorrelation graph both shows a trailing phenomenon and truncates to 0 after the q-th order, then select the autoregressive moving average model as the initial modeling model; p and q are both the orders of the initial modeling model; use the least squares method or the maximum likelihood estimation method to estimate the parameters of the initial modeling model and obtain the residual series; check whether the residual series is non-autocorrelated, has a mean of 0, and a constant variance; if the residual series is non-autocorrelated, has a mean of 0, and a constant variance, then record this residual series as the conditional series; Based on the conditional series, construct a data model for the corresponding grid area. The expression of the data model is: σ(t) 2 = a0 + ∑(s(u) × |∈(t - u)| x1 ) + ∑(e(v) × σ(t - v) 2 ) + ρ × D_t; where Among them, σ(t) 2 is the conditional variance at time t, a0 is a constant term greater than 0; s(u) is the autoregressive term coefficient; x1 is the exponential coefficient; e(v) is the generalized autoregressive term coefficient; σ(t - v) 2 is the conditional variance at the past time t - v; ∈(t - u) represents the value of the residual term of the time series at time t - u; u is the time lag order of the autoregressive term, v is the time lag order of the generalized autoregressive term; D_t is a 0-1 dummy variable; ρ is the coefficient of the dummy variable.

9. The digital engineering implementation dynamic monitoring system based on the Internet of Things according to claim 8, characterized in that The method of obtaining the optimal progress route includes: Regard the grid areas divided in the engineering site as nodes in an undirected graph. For adjacent grid areas, connect an undirected edge between them, and set the weight of each undirected edge as the difference in the conditional variances of the two adjacent grid areas at the future T time point; Select the starting point and the ending point in the grid area. If there are multiple starting points / ending points, connect them to the adjacent grid areas with undirected edges with a weight of 0; Run the shortest path algorithm on the undirected graph, with the starting point as the source point, find the shortest path from the starting point to the ending point, and the sequence of nodes on the shortest path constitutes the preliminary optimal progress route for the project implementation; perform smoothing processing on the preliminary optimal progress route to obtain the optimal progress route.

10. The dynamic monitoring method for digital engineering implementation based on the Internet of Things is realized based on the dynamic monitoring system for digital engineering implementation based on the Internet of Things described in any one of claims 1 to 9, and is characterized in that, Include: Step 1: Dynamically deploy several Internet of Things sensor nodes, and use the Internet of Things sensor nodes to collect the real-time data of the engineering site; Step 2: Integrate the collected real-time data into a data stream, and transmit it to the cloud server through a wireless communication network. On the cloud server, perform real-time segmentation on the incoming data stream to obtain relevant data groups; Step 3: Divide the project site into several grid areas, and perform data modeling on each grid area based on the relevant data sets to obtain the data model within the grid area; Step 4: Based on the data model within the grid area, use the shortest path algorithm to calculate the optimal progress route for project implementation; according to the optimal progress route, dynamically adjust the project implementation plan, and feedback the adjusted project implementation plan to the on-site construction guidance terminal.

Images