Big data-based construction engineering intelligent monitoring and safety early warning method and system

Abstract

The application provides a construction engineering intelligent monitoring and safety early warning method and system based on big data, relates to the technical field of engineering monitoring, and is used for improving the accuracy of construction engineering safety monitoring; the method mainly comprises the following steps: a regional edge calculation node determines a regional feature vector according to regional data of a corresponding construction region; and the regional feature vector is transmitted to a central calculation node; the central calculation node maps the regional feature vector of the construction region to a node corresponding to a construction graph structure of a target construction engineering; different nodes in the construction graph structure correspond to different construction regions; the central calculation node calculates a risk link and a corresponding risk level of the target construction engineering according to the construction graph structure of the target construction engineering; the risk link at least comprises one risk node; the central calculation node determines safety early warning information according to the level of the risk link and the corresponding risk level, and controls the output of the safety early warning information.

Description

Technical Field

[0001] This application relates to the field of engineering monitoring technology, and in particular to a method and system for intelligent monitoring and safety early warning of construction projects based on big data. Background Technology

[0002] Safety monitoring of construction projects aims to promptly detect abnormal changes during construction and prevent major safety accidents such as collapses, overturns, and falls. By monitoring structural stress, deformation, and environmental factors in real time, measures such as reinforcement, work stoppage, or evacuation can be taken at the early stage of a risk, protecting the lives of on-site workers, reducing property damage and project delays, and ensuring the smooth progress of the construction project.

[0003] Currently, inclinometers and axial force gauges are installed around the foundation pit and on the supporting structure of construction projects. Automated data acquisition devices measure the deep horizontal displacement of the soil and the stress on the supporting structure at regular intervals. The system automatically issues an alarm when the displacement or axial force exceeds a set warning value. However, the connecting cables of the inclinometers and axial force gauges, as well as the sensors themselves, are highly susceptible to damage during construction. Once damaged, the monitoring point becomes completely ineffective, resulting in low accuracy in safety monitoring of construction projects. Summary of the Invention

[0004] This application provides a method and system for intelligent monitoring and safety early warning of construction projects based on big data, which can improve the effectiveness and accuracy of safety monitoring of construction projects.

[0005] This invention provides a method for intelligent monitoring and safety early warning of construction projects based on big data. The method includes: The regional edge computing node acquires regional data of the corresponding construction area in the target construction project. The regional data includes visual image data, sensor data, equipment operation data, and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The edge computing node determines a regional feature vector based on the regional data of the corresponding construction area. The regional feature vector includes a regional data vector, risk prediction results, and regional information entropy. The node then transmits the regional feature vector to the central computing node. The central computing node maps the regional feature vector of the construction area to the node corresponding to the construction graph structure of the target construction project, where different nodes in the construction graph structure correspond to different construction areas; The central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project, wherein the risk link includes at least one risk node; The central computing node determines security warning information based on the level of the risk link and its corresponding risk level, and controls the output of the security warning information.

[0006] In an optional embodiment provided by the present invention, the regional edge computing node determines the regional feature vector based on the regional data of the corresponding construction area, including: The regional edge computing node determines the regional data vector based on the regional data of the corresponding construction area; The regional data vector is input into the engineering risk prediction model to obtain the risk prediction result, which includes the prediction probability corresponding to the prediction label set. The regional information entropy is determined based on the risk prediction results. The regional feature vector is obtained by combining the regional data vector, risk prediction results, and regional information entropy.

[0007] In an optional embodiment provided by the present invention, inputting the regional data vector into the engineering risk prediction model to obtain the risk prediction result includes: The expansion rate is determined based on the regional information entropy of the area constructed in the previous moment. The historical regional data vector of the construction area at the corresponding time point is obtained through the expansion rate. The risk prediction result is obtained by inputting the regional data vector at the current time point and the historical regional data vector into the engineering risk prediction model.

[0008] In an optional embodiment of the present invention, determining the expansion rate based on the regional information entropy of the constructed region at the previous moment includes: Obtain the regional information entropy of the construction area at the previous moment; If the regional information entropy of the area constructed at the previous moment is less than the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 1; If the regional information entropy of the area constructed at the previous moment is less than the second value and greater than or equal to the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 2. If the regional information entropy of the area constructed at the previous moment is greater than or equal to the second value, then the regional information entropy of the area constructed at the previous moment is determined to be 4.

[0009] In an optional embodiment of the present invention, the step of inputting the current time point regional data vector and the historical regional data vector into the engineering risk prediction model to obtain the risk prediction result includes: The current time point's regional data vector and the historical regional data vector are input into the engineering risk prediction model to obtain the initial risk prediction result; The initial risk prediction result is updated based on the current moving average error calculated by the engineering risk prediction model to obtain the final risk prediction result.

[0010] In an optional embodiment provided by the present invention, the central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project, including: Obtain the edge weights with connectivity relationships from the construction diagram structure of the target construction project; For each node in the construction diagram structure of the target construction project, the node iterative risk value is calculated based on the risk prediction results of the nodes and edge weights that have a connection with it. The risk links and corresponding risk levels of the target construction project are determined based on the iterative risk values ​​and the construction diagram structure of the target construction project.

[0011] In an optional embodiment of the present invention, the step of calculating the node iterative risk value for each node in the construction diagram structure of the target construction project based on the risk prediction results of the nodes and edge weights with which it has a connection relationship includes: In an optional embodiment of the present invention, determining the risk chain and corresponding risk level of the target construction project based on the iterative risk value and the construction diagram structure of the target construction project includes: Nodes with risk values ​​greater than preset risk values ​​are identified as target nodes, and the propagation risk contribution value of the predecessor nodes of the target nodes is calculated. The risk link is determined based on the propagation risk contribution value, and the risk level of the risk link is calculated based on the risk value of the nodes in the risk link.

[0012] This invention provides a big data-based intelligent monitoring and safety early warning system for construction projects, the system comprising: The regional edge computing node is used to acquire regional data of the corresponding construction area in the target construction project. The regional data includes visual image data, sensor data, equipment operation data and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The regional edge computing node is also used to determine a regional feature vector based on the regional data of the corresponding construction area. The regional feature vector includes a regional data vector, risk prediction results, and regional information entropy. The node then transmits the regional feature vector to the central computing node. The central computing node is used to map the regional feature vector of the construction area to the node corresponding to the construction map structure of the target construction project, wherein different nodes in the construction map structure correspond to different construction areas; The central computing node is used to calculate the risk links and corresponding risk levels of the target construction project based on the construction drawing structure of the target construction project, wherein the risk link includes at least one risk node; The central computing node is used to determine security warning information based on the hierarchy of the risk link and its corresponding risk level, and to control the output of the security warning information.

[0013] This invention provides a method and system for intelligent monitoring and safety early warning of construction projects based on big data. Regional edge computing nodes acquire regional data of the corresponding construction area in the target construction project. This regional data includes visual image data, sensor data, equipment operation data, and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The regional edge computing nodes determine regional feature vectors based on the regional data of the corresponding construction area. The regional feature vectors include regional data vectors, risk prediction results, and regional information entropy. These regional feature vectors are then transmitted to a central computing node. The central computing node maps the regional feature vectors of the construction areas to nodes corresponding to the construction graph structure of the target construction project. Different nodes in the construction graph structure correspond to different construction areas. Based on the construction graph structure of the target construction project, the central computing node calculates the risk links and corresponding risk levels of the target construction project. Each risk link includes at least one risk node. The central computing node determines safety early warning information based on the hierarchy of the risk links and their corresponding risk levels, and controls the output of the safety early warning information. This application divides large-scale construction projects into multiple areas and deploys edge computing nodes. The edge computing nodes process local sensor, image, equipment, and environmental data, extract lightweight regional feature vectors (including risk prediction and uncertainty entropy values), and upload them to the central computing node. This significantly reduces data transmission and the computational pressure on the central computing node, improving real-time performance. The central computing node maps each area to a node based on the construction graph structure and uses a risk propagation iterative algorithm to quantify risk links and levels, enabling clear tracing of risk sources and diffusion paths. Thus, this application improves the accuracy of construction project safety monitoring. Attached Figure Description

[0014] Figure 1 The following is a flowchart illustrating the execution of a big data-based intelligent monitoring and safety early warning method for construction projects, provided for this application. Figure 2 This application provides a structural diagram of a construction project intelligent monitoring and safety early warning system based on big data. Detailed Implementation

[0015] To better understand the above technical solutions, the technical solutions of the embodiments of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of this application and the specific features in the embodiments are detailed descriptions of the technical solutions of the embodiments of this application, rather than limitations on the technical solutions of this application. In the absence of conflict, the embodiments of this application and the technical features in the embodiments can be combined with each other.

[0016] Please see Figure 1 The image shows an embodiment of the present invention providing a method for intelligent monitoring and safety early warning of construction projects based on big data. The execution flow of this method is as follows: S101, the regional edge computing node obtains the regional data of the corresponding construction area in the target construction project.

[0017] The regional data includes visual image data, sensor data, equipment operation data, and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. Each edge computing node completes the acquisition of regional data for the corresponding construction area and the calculation of risk prediction results for the construction area, thereby reducing the computing pressure on the central computing node.

[0018] Specifically, in this embodiment, sensors can be deployed at key locations in the foundation pit and main structure. These sensors can include displacement data (such as horizontal displacement of the top of the foundation pit support piles and floor slab deflection), crack data (such as changes in the width of cracks on the wall or concrete surface), and tilt data (such as the tilt angle of the support column, tower body, and formwork support). At the same time, multiple sets of binocular vision cameras are deployed at high points covering all construction areas to obtain visual image data, such as images of cracks, seepage, and local collapse on the foundation pit slope, as well as visual records of water gushing from the bottom of the deep foundation pit and settlement cracks on the surrounding ground.

[0019] Equipment operation data can include: working status parameters of large machinery (such as working status parameters of tower cranes, construction hoists, excavators, etc.), equipment positioning and trajectory (such as GPS / BeiDou positioning data, used to determine whether large equipment has entered a dangerous area or is operating beyond its scope); environmental data can include: real-time rainfall, wind speed, temperature, humidity (affecting construction safety, such as prohibiting tower crane operation in strong winds), groundwater level changes, soil moisture content, and ground temperature (affecting the stability of the foundation pit and the safety of the slope).

[0020] S102, the regional edge computing node determines the regional feature vector based on the regional data of the corresponding construction area, and transmits the regional feature vector to the central computing node.

[0021] The regional feature vector includes a regional data vector, risk prediction results, and regional information entropy. In this embodiment, the regional data vector is a set of low-dimensional features (such as mean strain, tilt angle change rate, key point displacement, and tower crane overload count) extracted from raw sensor data, visual images, equipment operation data, and environmental data. This compresses massive amounts of raw data into key features, retaining the core information needed to assess risk while significantly reducing transmission volume and computational burden. The risk prediction results can be used by a time-series model to predict the probability of danger occurring in the region within a short future timeframe (e.g., 3-5 seconds). Incorporating risk prediction results allows for early identification of risk trends. The regional information entropy is an uncertainty measure calculated based on the risk prediction results, reflecting the model's degree of confidence in the region's state. A higher entropy value (closer to 1) indicates greater hesitation in the model; a lower entropy value indicates a more certain judgment.

[0022] In one optional embodiment provided in this application, the regional edge computing node determines the regional feature vector based on the regional data of the corresponding construction area, including: S1021, the regional edge computing node determines the regional data vector based on the regional data of the corresponding construction area.

[0023] Specifically, the regional data corresponding to the construction area is preprocessed, which involves performing operations such as feature extraction and data normalization on visual image data, sensor data, equipment operation data, and environmental data. Then, the data vectors corresponding to the visual image data, sensor data, equipment operation data, and environmental data are combined to obtain the regional data vector.

[0024] S1022, The regional data vector is input into the engineering risk prediction model to obtain the risk prediction result.

[0025] The risk prediction result includes the predicted probability corresponding to the predicted label set. Specifically, inputting the regional data vector into the engineering risk prediction model to obtain the risk prediction result includes: determining the inflation rate based on the regional information entropy of the construction area at the previous time; obtaining the historical regional data vector of the construction area at the corresponding time point through the inflation rate; and inputting the regional data vector of the current time point and the historical regional data vector into the engineering risk prediction model to obtain the risk prediction result. The inflation rate is used to control the sampling stride of the convolutional kernel on the input sequence.

[0026] In this embodiment, determining the expansion rate based on the regional information entropy of the construction area at the previous moment includes: obtaining the regional information entropy of the construction area at the previous moment; if the regional information entropy of the construction area at the previous moment is less than a first value, then determining the regional information entropy of the construction area at the previous moment to be 1; if the regional information entropy of the construction area at the previous moment is less than a second value but greater than or equal to the first value, then determining the regional information entropy of the construction area at the previous moment to be 2; if the regional information entropy of the construction area at the previous moment is greater than or equal to the second value, then determining the regional information entropy of the construction area at the previous moment to be 4.

[0027] Specifically, the expansion rate can be determined using the following formula;

[0028] Where r is the expansion rate, expansion rate If there is no historical entropy at the initial moment, the default value is r=1. ， The regional information entropy of the area built in the previous moment.

[0029] The dilation rate determines the sampling interval of the convolutional kernel on the time axis when constructing the input sequence. Let the current time be t, and L points from the past need to be sampled, but not consecutively; instead, they are sampled at equal intervals according to the dilation rate r.

[0030] Starting from the current moment, historical data points are taken at intervals of r, for a total of L points. For example, if r=2, L=3, and the current t=10 seconds, then the data vectors for the regions at times 10s, 8s, and 6s are taken. If r=1, then the data vectors for the regions at times 10s, 9s, and 8s are taken.

[0031] In this embodiment, the step of inputting the current time point regional data vector and the historical regional data vector into the engineering risk prediction model to obtain the risk prediction result includes: inputting the current time point regional data vector and the historical regional data vector into the engineering risk prediction model to obtain an initial risk prediction result; and updating the initial risk prediction result based on the current moving average error calculated by the engineering risk prediction model to obtain the final risk prediction result.

[0032] At each time step, when the true risk value is obtained Then, calculate the current prediction error. Then update the moving average error. ;in, This represents the true risk value. This is the initial risk prediction result. The moving average error at the current time point, initial value =0, This represents the moving average error at the previous time point. The forgetting factor, ranging from 0.9 to 0.99, is used to control the degree of retention of historical errors.

[0033] S1023, determine the regional information entropy based on the risk prediction results.

[0034] S1024, The regional feature vector is obtained by combining the regional data vector, the risk prediction result, and the regional information entropy.

[0035] In this embodiment, the regional data vector, risk prediction results, and regional information entropy are combined into a regional feature vector. On the one hand, this can effectively reduce false alarms and missed alarms caused by instantaneous noise or sensor anomalies, and improve the reliability of early warning. On the other hand, it provides the central node with differentiated regional risk characteristics, enabling it to dynamically adjust the sampling strategy and attention allocation according to the entropy value, thereby achieving more accurate and adaptive safety monitoring and resource scheduling in large-scale construction projects.

[0036] S103, the central computing node maps the regional feature vector of the construction area to the node corresponding to the construction diagram structure of the target construction project.

[0037] In this embodiment, different nodes in the construction graph structure correspond to different construction areas. The central computing node performs one-to-one matching based on the pre-established construction graph structure (each node corresponds to a pre-divided construction area) using area identifiers (such as area IDs or location codes): when it receives an area feature vector uploaded from an edge computing node of a certain area, the central computing node directly assigns the area ID carried in the vector to the corresponding numbered node in the construction graph structure as the current state attribute of that node, thereby completing the mapping from feature vectors to graph nodes.

[0038] S104, the central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project.

[0039] The risk link includes at least one risk node. In an optional embodiment provided in this application, the central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project, including: S1041, Obtain the edge weights with connection relationships from the construction diagram structure of the target construction project.

[0040] Each directed edge in the construction diagram structure weight This represents the intensity of the impact propagation on region j when a risk occurs in region i. Its initial weights can be determined through engineering experience calibration, historical data back-calculation, and spatial distance. Furthermore, upon completion of each construction stage (e.g., after each floor slab is poured), the weights of the affected edges are recalculated based on the latest spatial distance or process logic.

[0041] S1042, For each node in the construction diagram structure of the target construction project, calculate the node iterative risk value based on the risk prediction results of the nodes and edge weights that have a connection relationship with it.

[0042] Specifically, for each node in the construction diagram structure of the target construction project, the iterative risk value of the node is calculated based on the risk prediction results of the nodes and edge weights that have a connection with it, including: The node iteration risk value is calculated using the following formula:

[0043] in, Let i be the risk value of node i after the (k+1)th iteration. For the risk prediction result of node i, This is a risk attenuation factor, set based on engineering experience, and is typically set to 0.3~0.6. Let i be the set of all predecessor nodes of node i, read from the construction graph structure. Let be the edge weight from node j to node i. Let be the risk value of node i after the k-th iteration.

[0044] Furthermore, after calculating the node iterative risk value based on the risk prediction results and edge weights with which it is connected, the method further includes: whether the absolute value of the difference between the risk value of node i after the (k+1)th iteration and the risk value of node i after the kth iteration is less than a preset value; if it is less than the preset value, then the risk value of the last iteration is taken as the risk value of node i. The preset value can be set according to actual needs.

[0045] S1043, Based on the iterative risk value and the construction diagram structure of the target construction project, determine the risk link and corresponding risk level of the target construction project.

[0046] Specifically, determining the risk link and corresponding risk level of the target construction project based on the iterative risk value and the construction diagram structure of the target construction project includes: identifying nodes with risk values ​​greater than a preset risk value as target nodes, calculating the propagation risk contribution value of the predecessor nodes of the target nodes; determining the risk link based on the propagation risk contribution value, and calculating the risk level of the risk link based on the risk values ​​of the nodes in the risk link.

[0047] In this embodiment, the propagation risk contribution value of the predecessor node of the target node can be calculated using the following formula:

[0048] in, Let $j$ be the contribution value of the propagation risk from the predecessor node $j$ to the target node $t$. If so, then j is considered a valid source of risk. Let be the edge weight from node j to node t. Let θ be the risk value of node j in the last iteration. For each node that meets the condition, add it to the link and continue backtracking the predecessor node of node j until no predecessor with a contribution value exceeding θ can be found or the initial risk source is reached.

[0049] For example, consider three regions: foundation pit A (rA=0.8), support B (rB=0.2), and adjacent road C (rC=0.1). Edges: A→B weight 0.7, B→C weight 0.6, A→C weight 0.2. Let α=0.5.

[0050] C A→B =0.5×0.7×0.8=0.28 C A→C =0.5 × 0.2 × 0.8 = 0.08 C B→C =0.5 × 0.6 × 0.2 = 0.06 A has edges pointing to B and C, and the contribution values ​​of 0.28 and 0.08 are both greater than θ (0.05), so A can propagate to B and C. Two initial paths are obtained: A→B and A→C. For the path A→B, check if B has any outgoing edges pointing to other nodes. B points to C, and the contribution value of 0.06 is greater than 0.05, so it can be extended to A→B→C. For the path A→C, C has no outgoing edges (or no effective propagation), so the path terminates. With the terminal node C as the endpoint, there are two different risk propagation paths: direct path: A→C; indirect path: A→B→C. Using the sum of the risk values ​​of all nodes in the link as the level index, the sum of the node risk values ​​S for A→B→C is calculated to be 1.1. Based on the range of its value (0.8 ≤ 0.05), the risk propagation risk is determined. S <1.2) The risk level of the risk link A→B→C is determined to be medium risk.

[0051] S105, the central computing node determines the security warning information based on the level of the risk link and its corresponding risk level, and controls the output of the security warning information.

[0052] In one optional embodiment provided in this application, the central computing node determines specific early warning information by using a preset rule mapping table, combined with the level of the risk link (i.e., the number of nodes in the link, reflecting the scope of risk propagation) and the risk level (low, medium, high, extremely high). For example: short link (1~2 nodes) + high risk → triggers a local audible and visual alarm, notifying the safety officer in that area to check immediately; long link (≥3 nodes) + medium risk → sends an SMS to the project manager, requesting an expansion of the patrol scope and increased monitoring density; any link + extremely high risk → automatically outputs a shutdown command and notifies the entire site to evacuate via the broadcast system.

[0053] This invention provides a method for intelligent monitoring and safety early warning of construction projects based on big data. Regional edge computing nodes acquire regional data of the corresponding construction area in the target construction project. This regional data includes visual image data, sensor data, equipment operation data, and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The regional edge computing nodes determine regional feature vectors based on the regional data of the corresponding construction area. The regional feature vectors include regional data vectors, risk prediction results, and regional information entropy. These regional feature vectors are then transmitted to a central computing node. The central computing node maps the regional feature vectors of the construction areas to nodes corresponding to the construction graph structure of the target construction project. Different nodes in the construction graph structure correspond to different construction areas. Based on the construction graph structure of the target construction project, the central computing node calculates the risk links and corresponding risk levels of the target construction project. Each risk link includes at least one risk node. The central computing node determines safety early warning information based on the hierarchy of the risk links and their corresponding risk levels, and controls the output of the safety early warning information. This application divides large-scale construction projects into multiple areas and deploys edge computing nodes. The edge computing nodes process local sensor, image, equipment, and environmental data, extract lightweight regional feature vectors, and upload them to the central computing node, which significantly reduces data transmission and computing pressure and improves real-time performance. The central computing node maps each area to a node based on the construction graph structure and uses a risk propagation iterative algorithm to quantify risk links and levels, which can clearly trace the source and spread path of risks. Thus, this application improves the accuracy of safety monitoring of construction projects.

[0054] In one embodiment, a construction project intelligent monitoring and safety early warning system based on big data is provided. For example... Figure 2 As shown, the system includes: a regional edge computing node 21 and a central computing node 22; The regional edge computing node 21 is used to acquire regional data of the corresponding construction area in the target construction project. The regional data includes visual image data, sensor data, equipment operation data and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The regional edge computing node 21 is also used to determine a regional feature vector based on the regional data of the corresponding construction area, wherein the regional feature vector includes a regional data vector, risk prediction results, and regional information entropy; and transmit the regional feature vector to the central computing node. The central computing node 22 is used to map the regional feature vector of the construction area to the node corresponding to the construction map structure of the target construction project, wherein different nodes in the construction map structure correspond to different construction areas; The central computing node 22 is used to calculate the risk links and corresponding risk levels of the target construction project based on the construction drawing structure of the target construction project, wherein the risk links include at least one risk node; The central computing node 22 is used to determine security warning information based on the level of the risk link and its corresponding risk level, and to control the output of the security warning information.

[0055] In an optional embodiment provided by the present invention, the region edge computing node 21 is specifically used for: The regional edge computing node determines the regional data vector based on the regional data of the corresponding construction area; The regional data vector is input into the engineering risk prediction model to obtain the risk prediction result, which includes the prediction probability corresponding to the prediction label set. The regional information entropy is determined based on the risk prediction results. The regional feature vector is obtained by combining the regional data vector, risk prediction results, and regional information entropy.

[0056] In an optional embodiment provided by the present invention, the region edge computing node 21 is specifically used for: The expansion rate is determined based on the regional information entropy of the area constructed in the previous moment. The historical regional data vector of the construction area at the corresponding time point is obtained through the expansion rate. The risk prediction result is obtained by inputting the regional data vector at the current time point and the historical regional data vector into the engineering risk prediction model.

[0057] In an optional embodiment provided by the present invention, the region edge computing node 21 is specifically used for: Obtain the regional information entropy of the construction area at the previous moment; If the regional information entropy of the area constructed at the previous moment is less than the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 1; If the regional information entropy of the area constructed at the previous moment is less than the second value and greater than or equal to the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 2. If the regional information entropy of the area constructed at the previous moment is greater than or equal to the second value, then the regional information entropy of the area constructed at the previous moment is determined to be 4.

[0058] In an optional embodiment provided by the present invention, the region edge computing node 21 is specifically used for: The current time point's regional data vector and the historical regional data vector are input into the engineering risk prediction model to obtain the initial risk prediction result; The initial risk prediction result is updated based on the current moving average error calculated by the engineering risk prediction model to obtain the final risk prediction result.

[0059] In an optional embodiment provided by the present invention, the central computing node 22 is specifically used for: Obtain the edge weights with connectivity relationships from the construction diagram structure of the target construction project; For each node in the construction diagram structure of the target construction project, the node iterative risk value is calculated based on the risk prediction results of the nodes and edge weights that have a connection with it. The risk links and corresponding risk levels of the target construction project are determined based on the iterative risk values ​​and the construction diagram structure of the target construction project.

[0060] In an optional embodiment provided by the present invention, the central computing node 22 is specifically used for: The node iteration risk value is calculated using the following formula:

[0061] in, Let i be the risk value of node i after the (k+1)th iteration. For the risk prediction result of node i, As a risk attenuation factor, Let i be the set of all predecessor nodes of node i. Let be the edge weight from node j to node i. Let be the risk value of node i after the k-th iteration.

[0062] In an optional embodiment provided by the present invention, the central computing node 22 is further configured to: Is the absolute value of the difference between the risk value of node i after the (k+1)th iteration and the risk value of node i after the kth iteration less than the preset value? If the risk value is less than the preset value, then the risk value of the last iteration will be used as the risk value of node i.

[0063] In an optional embodiment provided by the present invention, the central computing node 22 is specifically used for: Nodes with risk values ​​greater than preset risk values ​​are identified as target nodes, and the propagation risk contribution value of the predecessor nodes of the target nodes is calculated. The risk link is determined based on the propagation risk contribution value, and the risk level of the risk link is calculated based on the risk value of the nodes in the risk link.

[0064] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0065] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0066] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for intelligent monitoring and safety early warning of construction projects based on big data, characterized in that, The method includes: The regional edge computing node acquires regional data of the corresponding construction area in the target construction project. The regional data includes visual image data, sensor data, equipment operation data, and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The edge computing node determines a regional feature vector based on the regional data of the corresponding construction area. The regional feature vector includes a regional data vector, risk prediction results, and regional information entropy. The node then transmits the regional feature vector to the central computing node. The central computing node maps the regional feature vector of the construction area to the node corresponding to the construction graph structure of the target construction project, where different nodes in the construction graph structure correspond to different construction areas; The central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project, wherein the risk link includes at least one risk node; The central computing node determines security warning information based on the level of the risk link and its corresponding risk level, and controls the output of the security warning information.

2. The method according to claim 1, characterized in that, The regional edge computing node determines the regional feature vector based on the regional data of the corresponding construction area, including: The regional edge computing node determines the regional data vector based on the regional data of the corresponding construction area; The regional data vector is input into the engineering risk prediction model to obtain the risk prediction result, which includes the prediction probability corresponding to the prediction label set. The regional information entropy is determined based on the risk prediction results. The regional feature vector is obtained by combining the regional data vector, risk prediction results, and regional information entropy.

3. The method according to claim 2, characterized in that, The risk prediction result is obtained by inputting the regional data vector into the engineering risk prediction model, including: The expansion rate is determined based on the regional information entropy of the area constructed in the previous moment. The historical regional data vector of the construction area at the corresponding time point is obtained through the expansion rate. The risk prediction result is obtained by inputting the regional data vector at the current time point and the historical regional data vector into the engineering risk prediction model.

4. The method according to claim 3, characterized in that, The step of determining the expansion rate based on the regional information entropy of the constructed area at the previous moment includes: Obtain the regional information entropy of the construction area at the previous moment; If the regional information entropy of the area constructed at the previous moment is less than the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 1; If the regional information entropy of the area constructed at the previous moment is less than the second value and greater than or equal to the first value, then the regional information entropy of the area constructed at the previous moment is determined to be 2. If the regional information entropy of the area constructed at the previous moment is greater than or equal to the second value, then the regional information entropy of the area constructed at the previous moment is determined to be 4.

5. The method according to claim 3, characterized in that, The step of inputting the current time point regional data vector and the historical regional data vector into the engineering risk prediction model to obtain the risk prediction result includes: The current time point's regional data vector and the historical regional data vector are input into the engineering risk prediction model to obtain the initial risk prediction result; The initial risk prediction result is updated based on the current moving average error calculated by the engineering risk prediction model to obtain the final risk prediction result.

6. The method according to any one of claims 1-5, characterized in that, The central computing node calculates the risk links and corresponding risk levels of the target construction project based on the construction diagram structure of the target construction project, including: Obtain the edge weights with connectivity relationships from the construction diagram structure of the target construction project; For each node in the construction diagram structure of the target construction project, the node iterative risk value is calculated based on the risk prediction results of the nodes and edge weights that have a connection with it. The risk links and corresponding risk levels of the target construction project are determined based on the iterative risk values ​​and the construction diagram structure of the target construction project.

7. The method according to claim 6, characterized in that, The step of determining the risk links and corresponding risk levels of the target construction project based on the iterative risk value and the construction diagram structure of the target construction project includes: Nodes with risk values ​​greater than preset risk values ​​are identified as target nodes, and the propagation risk contribution value of the predecessor nodes of the target nodes is calculated. The risk link is determined based on the propagation risk contribution value, and the risk level of the risk link is calculated based on the risk value of the nodes in the risk link.

8. A construction project intelligent monitoring and safety early warning system based on big data, characterized in that, The system includes: The regional edge computing node is used to acquire regional data of the corresponding construction area in the target construction project. The regional data includes visual image data, sensor data, equipment operation data and environmental data. The target construction project is divided into multiple construction areas, and different construction areas correspond to different regional edge computing nodes. The regional edge computing node is also used to determine a regional feature vector based on the regional data of the corresponding construction area. The regional feature vector includes a regional data vector, risk prediction results, and regional information entropy. The node then transmits the regional feature vector to the central computing node. The central computing node is used to map the regional feature vector of the construction area to the node corresponding to the construction map structure of the target construction project, wherein different nodes in the construction map structure correspond to different construction areas; The central computing node is used to calculate the risk links and corresponding risk levels of the target construction project based on the construction drawing structure of the target construction project, wherein the risk link includes at least one risk node; The central computing node is used to determine security warning information based on the hierarchy of the risk link and its corresponding risk level, and to control the output of the security warning information.

Images