Building construction safety monitoring method and system based on sensor

Abstract

The invention relates to the technical field of intelligent safety monitoring, and discloses a building construction safety monitoring method and system based on a sensor, and the method comprises the steps: carrying out the structural stress deployment of a building construction region, obtaining a dynamic twin scene, drawing a structural deformation curve of the building construction region, and carrying out the steel structure safety monitoring of the building construction region. Obtaining a first safety monitoring result; performing mixed gas decoupling on the building construction area to obtain a single gas characteristic spectral line, analyzing a single gas concentration value of the multi-gas area, and performing gas safety monitoring on the building construction area to obtain a second safety monitoring result; performing hot spot trajectory tracking on construction equipment in the building construction area to obtain a hot spot evolution trajectory, constructing a thermodynamic diagram of the construction equipment, and performing equipment overheating safety monitoring on the construction equipment to obtain a third safety monitoring result; and constructing a safety monitoring report. The safety monitoring comprehensiveness of the building construction area can be improved.

Description

TECHNICAL FIELD

[0001] The application relates to a sensor-based building construction safety monitoring method and system, and to the technical field of intelligent safety monitoring. BACKGROUND

[0002] Building construction safety is a key link for ensuring the smooth progress of a project and protecting the safety of personnel and property, and plays a very important role in various construction projects. Whether it is the construction of high-rise buildings or the construction of bridges and roads, real-time and accurate safety monitoring can timely discover potential risks and prevent safety accidents, thereby becoming a core requirement for improving construction management level and ensuring engineering quality.

[0003] Currently, the monitoring of building construction safety is mainly manual, and a safety inspector uses some simple instruments to monitor and collect data on the local environment or structure of a building construction area, for example, using a range finder to monitor the depth of a foundation pit. However, with the continuous expansion of the scale of building projects and the increasing complexity of the construction environment, the device structure, construction equipment and safety influencing factors of the construction site are changing, which leads to the fact that the traditional safety monitoring method is not comprehensive enough for the safety monitoring of many building construction areas. SUMMARY

[0004] The application provides a sensor-based building construction safety monitoring method and system, which mainly aims to improve the comprehensiveness of safety monitoring of a building construction area.

[0005] To achieve the above-mentioned purpose, the application provides a sensor-based building construction safety monitoring method, which comprises the following steps: Collecting multi-sensor data of a building construction area, wherein the multi-sensor data comprises structural sensor data, infrared sensor data and gas sensor data, and the structural sensor data comprises steel structure vibration data, steel structure inclination data and steel structure stress data; Converting the structural sensor data into a joint embedding vector, using the joint embedding vector to perform structural stress deployment on the building construction area, obtaining a dynamic twin scene, using the dynamic twin scene to draw a structural deformation curve of the building construction area, using the structural deformation curve to perform steel structure safety monitoring on the building construction area, and obtaining a first safety monitoring result; Based on the gas sensor data, performing mixed gas decoupling on the building construction area, obtaining a single gas characteristic spectrum line, using the single gas characteristic spectrum line to analyze the single gas concentration value of the multi-gas area, using the single gas concentration value to perform gas safety monitoring on the building construction area, and obtaining a second safety monitoring result; The infrared sensing data is used for tracking a hot spot trajectory of a construction equipment in the construction area, to obtain a hot spot evolution trajectory, and a heat map of the construction equipment is constructed based on the hot spot evolution trajectory, and the construction equipment is monitored for overheating safety based on the heat map, to obtain a third safety monitoring result; Based on the first safety monitoring result, the second safety monitoring result and the third safety monitoring result, a safety monitoring report of the construction area is constructed.

[0006] Optionally, the structural sensing data is converted into a joint embedding vector, including: Three-dimensional features of the structural sensing data are extracted, to obtain a vibration vector, an inclination vector and a stress vector; The vibration vector is processed by wavelet packet transform, to obtain a wavelet transform vector; The inclination vector is processed by quaternion differential, to obtain a four-dimensional angle vector; The stress vector is processed by principal stress analysis, to obtain a joint stress vector; The wavelet transform vector, the four-dimensional angle vector and the joint stress vector are mapped across modalities, to obtain a preliminary fusion feature vector; The preliminary fusion feature vector is subjected to spatial semantic constraints, to obtain a joint embedding vector.

[0007] Optionally, the joint embedding vector is used for structural stress deployment of the construction area, to obtain a dynamic twin scene, including: The joint embedding vector is subjected to mechanical parameter decoupling, to obtain a structured mechanical parameter set; The structured mechanical parameter set is used to analyze a transient mechanical field of the construction area; According to the transient mechanical field, a stress-geometry coupling model of the construction area is constructed; The stress-geometry coupling model is visualized and rendered, to obtain a dynamic twin scene.

[0008] Optionally, the dynamic twin scene is used to draw a structural deformation curve of the construction area, including: The dynamic twin scene is subjected to stress gradient clustering processing, to identify a high stress node group in the dynamic twin scene; A displacement field matrix of stress nodes in the high stress node group is constructed; Based on the displacement field matrix, a node equivalent deformation variable in the dynamic twin scene is calculated; Based on the node equivalent deformation variable, a structural deformation curve of the construction area is drawn.

[0009] Optionally, the steel structure safety monitoring is performed on the building construction area based on the structure deformation curve to obtain a first safety monitoring result, including: calculating a deformation rate of the structure deformation curve; identifying a steel structure mutation point in the building construction area based on the deformation rate to obtain an abnormal point set; identifying a steel structure abnormal pattern of the building construction area according to the abnormal point set; performing steel structure safety monitoring on the building construction area based on the steel structure abnormal pattern to obtain a first safety monitoring result.

[0010] Optionally, the calculating a deformation rate of the structure deformation curve includes: performing sliding window processing on the structure deformation curve to obtain a sampling curve; calculating a deformation rate of the sampling curve using the following formula: ; wherein, denotes the deformation rate of the sampling curve, denotes an i-th sliding window sampling timestamp, n denotes a sliding window radius, and k denotes a relative position index in the sliding window, denotes expanding a boundary position backward by n time steps, denotes expanding a boundary position forward by n time steps, denotes a Gaussian weight of k, denotes a sampling interval, denotes a deformation variable at time t, denotes a deformation variable at time t.

[0011] Optionally, based on the gas sensing data, mixed gas decoupling is performed on the building construction area to obtain single gas characteristic spectral lines, including: performing data standardization processing on the gas sensing data to obtain standardized gas data; constructing a mixed gas feature matrix of the building construction area using the standardized gas data; identifying absorption peak intensity and wavelength of mixed gas in the building construction area using the mixed gas feature matrix; based on the absorption peak intensity and the wavelength, calculating a spectral signal intensity of the mixed gas using the following formula: ; wherein, the spectral signal intensity, N denotes the number of mixed gases, an absorption peak intensity of the jth gas in the mixed gas, a wavelength of the mixed gas, a central wavelength representing j, a standard deviation representing j; decoupling the mixed gas based on the spectral signal intensity, to obtain a single gas characteristic spectrum.

[0012] Optionally, using the single gas characteristic spectrum, analyzing the single gas concentration value of the multi-gas region, comprising: baseline correction is performed on the single gas characteristic spectrum to obtain a pure absorption peak spectrum; peak value determination is performed on the pure absorption peak spectrum to obtain an absorption peak intensity vector at a specific wavelength in the pure absorption peak spectrum; based on the absorption peak intensity vector, using the Lambert-Beer calibration model, identifying the initial concentration value of the single gas in the multi-gas region; cross-interference correction is performed on the initial concentration value of the single gas to obtain the single gas concentration value.

[0013] Optionally, using the infrared sensing data, tracking the hot spot trajectory of the construction equipment in the construction area to obtain a hot spot evolution trajectory, comprising: using the infrared sensing data, constructing a temperature matrix of the construction area; based on the temperature matrix, identifying the hot spot of the construction equipment in the construction area; performing multi-frame trajectory correlation on the hot spot to obtain a hot spot evolution trajectory.

[0014] In order to solve the above problems, the application also provides a kind of based on sensor's construction safety monitoring system of construction, the system comprises: data acquisition module, for collecting the multi-sensor data of construction area, wherein the multi-sensor data contains structural sensing data, infrared sensing data and gas sensing data, the structural sensing data contains steel structure vibration data, steel structure inclination data and steel structure stress data; the first monitoring module is used to convert the structural sensing data into a joint embedding vector, and the joint embedding vector is used to deploy the structural stress of the construction area to obtain a dynamic twin scene, and the dynamic twin scene is used to draw a structural deformation curve of the construction area, and the structural deformation curve is used to monitor the steel structure safety of the construction area to obtain a first safety monitoring result; The second monitoring module is configured to decouple mixed gas in the construction area based on the gas sensing data to obtain single-gas characteristic spectral lines, analyze single-gas concentration values of the multi-gas area by using the single-gas characteristic spectral lines, and perform gas safety monitoring on the construction area by using the single-gas concentration values to obtain a second safety monitoring result. The third monitoring module is configured to track a hot spot trajectory of the construction equipment in the construction area by using the infrared sensing data to obtain a hot spot evolution trajectory, construct a heat map of the construction equipment based on the hot spot evolution trajectory, and perform equipment overheating safety monitoring on the construction equipment based on the heat map to obtain a third safety monitoring result. The safety monitoring report module is configured to construct a safety monitoring report of the construction area based on the first safety monitoring result, the second safety monitoring result, and the third safety monitoring result.

[0015] The present application first collects the vibration frequency, inclination angle and internal stress data of the steel structure by deploying vibration, inclination and stress sensors, obtains the thermal radiation energy of the equipment surface by using the infrared sensor, and monitors the mixed gas concentration by using the gas sensor, so as to realize multi-dimensional data coverage of the construction area structure state, equipment temperature and gas composition, and provide real-time basic data support for subsequent analysis. Further, the present application helps users to analyze the deformation trend, rate and overrun of the steel structure in the construction process by combining the three steps of embedding vector generation, dynamically constructing a twin scene and drawing a structure deformation curve, so as to timely discover potential risks such as stress concentration and abnormal deformation of components. Further, the present application can compare the single-gas concentration value measured in the construction area with the pre-set gas safety threshold (such as the lower limit of methane explosion and the critical value of carbon monoxide poisoning) by using the two steps of single-gas characteristic spectral line extraction and single-gas concentration calculation, and issues a safety warning if the concentration exceeds the threshold, or determines safety if it does not exceed the threshold. Further, the present application analyzes the position, range and temperature value of the high-temperature area in the equipment temperature distribution cloud map by using the two methods of hot spot evolution trajectory analysis and heat map construction, and determines whether the construction equipment has an overheating fault. Further, the present application can integrate and summarize the scattered risk information of the steel structure safety, gas safety and equipment overheating in the construction area into a structured report form to present the overall safety situation, and provide comprehensive and accurate decision-making basis for users. Therefore, the present application can improve the comprehensiveness of the safety monitoring of the construction area. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A flowchart of a sensor-based construction safety monitoring method according to an embodiment of the present application is shown in the figure. Figure 2 A module diagram of a sensor-based construction safety monitoring method according to an embodiment of the present application is shown in the figure.

[0017] The objectives, functional characteristics and advantages of the present application will be further explained in conjunction with the embodiments, with reference to the accompanying drawings. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein merely exemplify the application and do not limit the application.

[0019] Embodiments of the present application provide a kind of based on sensor's building construction safety monitoring method.The execution subject of the kind of based on sensor's building construction safety monitoring method includes but is not limited to server, terminal and at least one of the electronic equipment that can be configured to execute the method provided in the present application, etc.It is said in other words, the kind of based on sensor's building construction safety monitoring method can be executed by software or hardware installed in terminal equipment or server equipment.The server includes but is not limited to: single server, server cluster, cloud server or cloud server cluster, etc.

[0020] Embodiment 1: Referring to Figure 1 As shown in the figure, a kind of based on sensor's building construction safety monitoring method provided by an embodiment of the present application is flow chart diagram.In the embodiment, the kind of based on sensor's building construction safety monitoring method includes: S1, the multi-sensor data of building construction area is collected, wherein the multi-sensor data includes structural sensing data, infrared sensing data and gas sensing data, and the structural sensing data includes steel structure vibration data, steel structure inclination data and steel structure stress data.

[0021] The multi-sensor data of building construction area can be collected by deploying multiple sensors to collect structural vibration, infrared hot spot and gas concentration data in the embodiment of the present application, which provides multi-dimensional real-time data support for construction safety monitoring.

[0022] The multi-sensor data includes structural sensing data, infrared sensing data and gas sensing data, and the structural sensing data includes steel structure vibration data, steel structure inclination data and steel structure stress data.

[0023] Further, the infrared sensing data refers to the thermal radiation energy data of the surface of the construction equipment collected by the infrared sensor, which can reflect the temperature distribution of each part of the equipment, such as the temperature abnormally high signal of the tower crane motor during operation, the gas sensing data refers to the mixed gas composition and concentration data of the construction area obtained by the gas sensor, such as the concentration value of harmful gases such as benzene and formaldehyde generated by paint volatilization during basement construction, the steel structure vibration data refers to the vibration frequency, amplitude and other parameters of the steel structure under the action of construction load or environment collected by the vibration sensor, such as the vibration amplitude change of the steel formwork during concrete pouring, the steel structure inclination data refers to the inclination angle and change rate data of the steel structure member measured by the inclination sensor, which can be used to monitor the verticality deviation of the scaffold or tower body, such as the small inclination amount of the steel structure column in high-rise construction, and the steel structure stress data refers to the internal stress distribution and change data of the steel structure measured by the stress sensor, such as the stress concentration of the steel box girder under load during bridge construction.

[0024] S2, convert the structure sensing data into a joint embedding vector, deploy the structure stress of the building construction area by using the joint embedding vector, obtain a dynamic twin scene, draw a structure deformation curve of the building construction area by using the dynamic twin scene, and monitor the safety of the steel structure of the building construction area by using the structure deformation curve, to obtain a first safety monitoring result.

[0025] The embodiment of the application can map the vibration, inclination and stress and other multi-dimensional data in the building foundation structure in the building construction area to a unified feature space by converting the structure sensing data into a joint embedding vector, which facilitates the analysis of the correlation between the steel structure and the data features.

[0026] The joint embedding vector is a unified feature vector formed by fusing the data features of the multi-dimensional structure sensing data such as vibration, inclination and stress of the steel structure.

[0027] As an embodiment of the application, converting the structure sensing data into a joint embedding vector comprises: extracting three-dimensional features of the structure sensing data to obtain a vibration vector, an inclination vector and a stress vector; performing wavelet packet transform processing on the vibration vector to obtain a wavelet transform vector; performing quaternion differential processing on the inclination vector to obtain a four-dimensional angle vector; performing principal stress analysis processing on the stress vector to obtain a joint stress vector; performing cross-modal mapping on the wavelet transform vector, the four-dimensional angle vector and the joint stress vector to obtain a preliminary fusion feature vector; The preliminary fusion feature vector is subjected to spatial semantic constraint to obtain a joint embedding vector.

[0028] The four-dimensional angle vector is a four-dimensional feature vector obtained by processing the inclination data through quaternion differential processing, and is used to represent the attitude change and rotational dynamics characteristics of the structure in space.

[0029] In a specific implementation, the vibration data in the structure sensing data can be subjected to root mean square value and peak factor calculation, and then the part of the 10-50Hz frequency band energy ratio is extracted to obtain a vibration vector, the pitch angle, roll angle and yaw angle in the inclination data are extracted to obtain an inclination vector, and the stress data is converted into a stress value to obtain a stress vector; the vibration vector is decomposed into a multi-band wavelet coefficient, the energy ratio of each frequency band is calculated and the feature vector is reconstructed to obtain a wavelet transform vector; the Euler angle in the inclination vector is converted into a quaternion representation, and then the attitude change rate of the quaternion is calculated through a quaternion differential equation to obtain a four-dimensional angle vector; the eigenvalues and eigenvectors of the stress vector are calculated based on the stress tensor to determine the direction and size of the principal stress, and a joint stress vector containing normal stress and shear stress is constructed; the wavelet transform vector, the four-dimensional angle vector and the joint stress vector are mapped into a multi-dimensional space (such as 512 dimensions) through a linear layer, and then the vector weight after mapping is calculated using a multi-head attention mechanism, and then the vector fusion is performed based on the weight to obtain a preliminary fusion feature vector; a contrast loss function can be designed through a deep neural network to force the feature vectors of the same working condition to be adjacent in space and the feature vectors of different working conditions to be far away, and after optimizing the semantic representation ability of the feature vectors, a joint embedding vector is obtained, for example, after network training, the distance between the joint embedding vectors of "steel structure normal loading" and "loose bolt" in the feature space increases from 1.2 to 2.8, and when the distance of the new vector is less than 2.0, it indicates that there may be an abnormality, which can be focused on at this time.

[0030] The embodiment of the present application can dynamically and accurately simulate the stress distribution and deformation of the building structure under the construction load by using the joint embedding vector to deploy the structure stress of the building construction area to obtain a dynamic twin scene, for example, in the hoisting construction of a high-rise steel structure, the stress concentration area and deformation trend of the steel member during hoisting can be intuitively presented based on the joint embedding vector, which helps construction personnel to timely find potential structural safety hazards.

[0031] The dynamic twin scene refers to mapping the real-time sensing data of the physical building construction area to a virtual three-dimensional space to construct a digital mirror scene that can be updated in real time and dynamically interacted.

[0032] As an embodiment of the present application, the joint embedding vector is used for structural stress deployment of the construction area, to obtain a dynamic twin scene, comprising: The joint embedding vector is mechanically decoupled to obtain a structured mechanical parameter set; The structured mechanical parameter set is used to analyze the transient mechanical field of the construction area; According to the transient mechanical field, a stress-geometry coupling model of the construction area is constructed; The stress-geometry coupling model is visualized and rendered to obtain a dynamic twin scene.

[0033] The structured mechanical parameter set refers to a standardized physical parameter set obtained by mechanically decoupling the joint embedding vector, which contains stress components, strain rates, and vibration frequencies, etc. The transient mechanical field refers to the spatial distribution field of stress, strain, displacement, etc. of the building structure under a certain transient working condition based on the structured mechanical parameters, which is usually represented in the form of a cloud chart or a tensor field to represent the dynamic stress state. The stress-geometry coupling model refers to a mechanical model that relates the stress state of the structure to the deformed geometry.

[0034] In the specific implementation process, the high-dimensional joint embedding vector can be decomposed into parameters that can directly drive mechanical calculation through inverse mapping algorithm, and physical quantities such as stress components, strain rates, and vibration modes are extracted. For example, when a large-span steel structure roof is lifted, the decoupling obtains x-tension stress 120 MPa, y-compression stress-50 MPa, and first-order vibration frequency 8.7 Hz, forming a structured parameter set. Based on the decoupled mechanical parameters, the stress, strain distribution and dynamic response of the structure under the current working condition are calculated by the finite element method (FEM), for example, when a subway station deep foundation pit is excavated, the transient maximum bending moment of the retaining pile is 280 kN·m, and the displacement of the soil behind the wall is 15 mm, forming a mechanical field cloud chart data. The mechanical field analysis results are associated with the structure geometry model, and the mutual influence of stress and geometry under large deformation (geometric nonlinearity) is considered, for example, when a high-rise building steel frame is constructed, the coupling model shows that the core tube angle column is shortened by 5 mm due to axial compression, resulting in an increase of 12% in beam end stress redistribution; the mechanical response (stress cloud chart, deformation animation) of the coupling model is fused with the BIM model, and a dynamic twin scene is generated through a real-time rendering engine.

[0035] Further, by using the dynamic twin scene, the embodiment of the present application can present the real-time deformation data of the building structure in the virtual scene as a visual curve, intuitively reflect the change trend of the structure in the construction process, such as displacement and inclination, and the like, for example, in the bridge cantilever pouring construction, by drawing the deflection change curve of the beam body during pouring of each segment, abnormal deformation can be found in time, and data support is provided for construction safety and quality control.

[0036] The structure deformation curve refers to a continuous visual curve formed by data fitting, and is used to intuitively display the deformation dynamic characteristics of the structure in the construction process.

[0037] As an embodiment of the present application, by using the dynamic twin scene, the structure deformation curve of the building construction area is drawn, including: Performing stress gradient clustering processing on the dynamic twin scene to identify a high stress node group in the dynamic twin scene; Constructing a displacement field matrix of the stress nodes in the high stress node group; Based on the displacement field matrix, calculating a node equivalent deformation variable in the dynamic twin scene; Based on the node equivalent deformation variable, drawing the structure deformation curve of the building construction area.

[0038] The high stress node group refers to a node set with similar stress gradient values and dense spatial distribution in the dynamic twin scene, which is identified by a stress gradient clustering algorithm (such as DBSCAN), the displacement field matrix refers to a multi-dimensional matrix recording the three-dimensional displacement (X / Y / Z direction) of the nodes, which is constructed by taking each node in the high stress node group as a row and different monitoring time as a column, and is used to quantitatively store the space-time data of the structure deformation, and the node equivalent deformation variable refers to a comprehensive deformation variable calculated by vector synthesis based on the three-dimensional displacement data of the nodes, which reflects the actual deformation degree (i.e. the size of the combined displacement) of the nodes in space.

[0039] In the implementation process, a density clustering algorithm such as DBSCAN can be used to calculate the stress gradient values of each node in the dynamic twin scene, and according to the set density threshold and neighborhood radius, nodes with similar stress gradients are divided into the same high stress node group, for example, in the construction of a large-span steel truss bridge, the stress data of the dynamic twin scene is processed by the DBSCAN algorithm: set the neighborhood radius ε = 5 MPa / m (stress gradient difference threshold), the minimum sample number MinPts = 3, and by calculating the Euclidean distance between the stress gradients of the nodes, 18 nodes (such as the lower chord nodes in the mid-span region) with gradient values of 12-15 MPa / m are divided into a high stress node group; the three-dimensional displacement coordinates of each node in the high stress node group at different times are obtained, and these displacement coordinates are arranged in order of node number and time sequence to construct a multi-dimensional displacement field matrix, for example, in the construction process of a steel box girder bridge, for high stress node groups such as support connections, a displacement field matrix containing the X, Y, and Z direction displacements of each node is constructed to record the displacement changes at different construction stages; for each node displacement data in the displacement field matrix, the equivalent deformation of the node is obtained by calculating the resultant displacement (i.e. the vector sum of the three-dimensional displacement), which can be calculated by the following formula: ; wherein, represents the equivalent deformation of the node, represents the x-axis displacement of the stress node, represents the -axis displacement of the stress node, represents the -axis displacement of the stress node.

[0040] Further, the equivalent deformation of each node can be plotted using a data visualization tool with the construction time or construction progress as the horizontal coordinate and the equivalent deformation of the node as the vertical coordinate, and a smooth structure deformation curve can be generated by a curve fitting algorithm.

[0041] The structure deformation curve can be used to monitor the safety of the steel structure in the building construction area, and the first safety monitoring result can help users analyze the deformation trend, rate and overrun of the steel structure during construction, so as to timely discover potential risks such as stress concentration and abnormal deformation of components, for example, in the construction of a high-rise steel structure building, by analyzing the perpendicularity deviation curve of the core tube steel column, if the slope of the curve suddenly increases at a certain time period, it indicates that the steel column may be tilted, and reinforcement measures can be taken immediately to avoid safety accidents.

[0042] As an embodiment of the present application, the structure deformation curve is used to monitor the safety of the steel structure in the building construction area, and the first safety monitoring result includes: calculating the deformation rate of the structure deformation curve; Based on the deformation rate, a steel structure mutation point in the construction area is identified to obtain an abnormal point set; According to the abnormal point set, a steel structure abnormal mode of the construction area is identified; Based on the steel structure abnormal mode, steel structure safety monitoring is performed on the construction area to obtain a first safety monitoring result.

[0043] The deformation rate refers to the change amount in unit time, that is, the first derivative of the deformation amount with respect to time, which reflects the speed of structure deformation and is used to evaluate the severity of deformation development trend. The abnormal point set refers to a set of time points that deviate from the normal deformation rule and are screened out by the deformation rate threshold. The structure state corresponding to these points may have safety hazards, which are usually manifested as sudden change or continuous overrun of the deformation rate. The steel structure abnormal mode refers to a typical safety hazard type that is matched after clustering analysis of the spatiotemporal characteristics of the abnormal point set and the corresponding construction conditions, such as support instability, load mutation, and welding defects.

[0044] In the specific implementation process, the abnormal point set can be obtained by setting a deformation rate threshold (such as 2 times the standard deviation under normal conditions, which needs to be set in combination with actual application). When the rate of a certain point exceeds the threshold and the rate difference between the previous and subsequent time points exceeds a set value, it is determined as a mutation point. For example, when a steel structure roof is hoisted, the deformation rate suddenly increases from 0.5 mm / day to 4.2 mm / day, and the second derivative reaches 3.7 mm / day². It is determined that the time point is a mutation point and is added to the abnormal point set. The time distribution, deformation amount, and corresponding construction process of the abnormal point set are clustered and analyzed to match a pre-set abnormal mode library (such as load mutation, support failure, material defect, etc.) for identification. For example, the abnormal point set shows that the deformation rate suddenly increases at a certain time period and corresponds to the steel structure welding process. After clustering, it is matched to the “welding thermal stress causing local deformation” mode, which is characterized by sudden increase of one-way deformation in a short time. According to the identified abnormal mode, the corresponding safety evaluation rule is called, and the safety monitoring conclusion and processing suggestion are generated in combination with the specification threshold, for example, after identifying the “temporary support settlement” abnormal mode, the support point displacement in the dynamic twin scene reaches 12 mm (> 10 mm threshold), and the first safety monitoring result determines that the support system is at risk of failure, and the temporary support should be immediately reinforced.

[0045] Further, as an optional embodiment of the present application, the calculation of the deformation rate of the structure deformation curve comprises: The structure deformation curve is subjected to sliding window processing to obtain a sampling curve; The deformation rate of the sampling curve is calculated using the following formula: ; Wherein, represents the deformation rate of the sampling curve, denotes the i-th sliding window sampling timestamp, n denotes the sliding window radius, k denotes the relative position index within the sliding window, denotes the boundary position is extended backward by n time steps, denotes the boundary position is extended forward by n time steps, denotes the Gaussian weight of k, denotes the sampling interval, denotes 1 the deformation variable at time t, denotes the deformation variable at time t.

[0046] It should be noted that the above deformation rate calculation formula performs local sampling on the structure deformation curve through a sliding window, expands n time steps forward and backward from the current time point to form an analysis window, assigns weights to each time point in the window using a Gaussian function (the closer to the current point, the higher the weight), and then calculates the weighted average rate of change of the deformation variable in the window through bidirectional difference. The core is to highlight the dominant role of the current time deformation through Gaussian weighting, while combining adjacent period data to suppress noise, and the bidirectional difference design takes into account the correlation of the deformation trend before and after, and the boundary constraint ensures that the window can still be effectively calculated at both ends of the curve.

[0047] Further, in actual application, for example, in steel structure construction monitoring, this method can quickly capture deformation rate mutation (such as deformation surge when the tower crane wall is removed), which can provide early warning compared to traditional single-point difference, and reduce false alarm rate caused by noise. By adjusting the window radius n and the Gaussian standard deviation, different construction stages (such as small n to improve sensitivity during hoisting period, and large n to enhance stability during maintenance period) can be adapted to achieve accurate quantification of the dynamic characteristics of the structure deformation.

[0048] S3, based on the gas sensing data, decoupling the mixed gas of the building construction area to obtain a single gas characteristic spectrum, using the single gas characteristic spectrum, analyzing the single gas concentration value of the multi-gas area, using the single gas concentration value, gas safety monitoring of the building construction area is carried out, and a second safety monitoring result is obtained.

[0049] The embodiment of the application can separate the unique spectral characteristics of each gas by decoupling the mixed gas of the building construction area based on the gas sensing data to accurately identify and quantitatively detect various gas components.

[0050] The single gas characteristic spectrum refers to the unique spectral absorption or emission characteristic curve that each gas has in a specific wavelength range.

[0051] As an embodiment of the present application, based on the gas sensing data, the construction area is decoupled for mixed gas to obtain single gas characteristic spectrum, including: The gas sensing data is subjected to data standardization processing to obtain standardized gas data; The standardized gas data is used to construct a mixed gas characteristic matrix of the construction area; The mixed gas characteristic matrix is used to identify the absorption peak intensity and wavelength of the mixed gas in the construction area; Based on the absorption peak intensity and the wavelength, the spectral signal intensity of the mixed gas is calculated by the following formula: ; Wherein, The spectral signal intensity, N represents the number of mixed gases, represents the absorption peak intensity of the jth gas in the mixed gas, The wavelength of the mixed gas, represents the central wavelength of j, represents the standard deviation of j; Based on the spectral signal intensity, the construction area is decoupled for mixed gas to obtain single gas characteristic spectrum.

[0052] Wherein, the mixed gas characteristic matrix refers to a two-dimensional data matrix constructed with time series as rows and gas sensing values at different wavelengths as columns, used to store the spectral response characteristics of the mixed gas in the construction area, the absorption peak intensity refers to the maximum absorption capacity of a single gas in the mixed gas at a characteristic wavelength, and the spectral signal intensity refers to the comprehensive light intensity value obtained by weighted summation of Gaussian functions considering the absorption contribution of each single gas in the mixed gas at different wavelengths.

[0053] In specific implementation, the Z-score standardization method can be used for data standardization processing of the gas sensing data; a two-dimensional matrix is constructed with time series as row index and sensing values of each gas at different wavelengths as column index; for a multi-channel sensor, the response values at different wavelengths at the same time are arranged as a row of the matrix, and the complete matrix is stacked in time sequence; local maximum value search is performed on each column (time series of a specific wavelength) of the characteristic matrix, a minimum peak height threshold (such as 0.3 times the standard deviation) and a minimum peak spacing (such as 2 sampling points) are set, and the true absorption peak is filtered out and the wavelength position and corresponding intensity value of each peak are recorded; the spectral signal intensity matrix is decomposed into the product of a single gas characteristic matrix S (wavelength x gas number) and a concentration weight matrix C (gas number x time), the elements of S and C are non-negative, the decomposition error is minimized through an iterative optimization algorithm (such as multiplication update rule), and the characteristic spectrum of each gas is obtained.

[0054] It should be noted that the formula for calculating the spectral signal intensity treats the spectral signal intensity of the mixed gas at a specific wavelength as the superposition of the absorption contributions of each individual gas, and simulates the shape (center wavelength) of the absorption peak of each gas using a Gaussian function. Determines location, standard deviation Characterizing peak width, to absorbance peak intensity This method uses the summation of Gaussian curves of each individual gas as weights to achieve a mathematical model of the mixed spectrum. Its core principle is the assumption that each gas absorbs independently and does not interfere with each other, reconstructing the overall spectral signal through linear superposition. In practical applications, such as in building construction gas monitoring, this formula can decompose the complex spectrum of a mixed gas into the characteristic contributions of each gas. For example, in scenarios where methane (3.3 μm) and carbon monoxide (4.6 μm) are present simultaneously, it can accurately separate the characteristic spectral lines of the two gases, reducing the methane concentration detection error from ±15% to ±5% (the specific error needs to be determined based on actual application testing). By adjusting... The parameters are adapted to the absorption peak width characteristics of different gases. For combustible gases with a concentration range of 0-10% LEL, the decoupling accuracy is improved by 30%, effectively identifying early leaks below 1% LEL, and providing more reliable data support for construction safety early warning.

[0055] Furthermore, by utilizing the single-gas characteristic spectral lines, the embodiments of the present invention can accurately analyze the concentration of each component in the mixed gas by analyzing the single-gas concentration values ​​in the multi-gas region, thereby promptly detecting potential risks due to excessively high concentrations of harmful gases.

[0056] As an embodiment of the present invention, the analysis of the single-gas concentration value in the multi-gas region using the single-gas characteristic spectral lines includes: Baseline correction was performed on the single-gas characteristic spectral lines to obtain pure absorption peak spectral lines; The peak value of the pure absorption peak spectrum is measured to obtain the absorption peak intensity vector at a specific wavelength in the pure absorption peak spectrum. Based on the absorption peak intensity vector, the initial concentration value of a single gas in the multi-gas region is identified using the Lambert-Beer calibration model. Cross-interference correction is performed on the initial concentration value of the single gas to obtain the concentration value of the single gas.

[0057] The pure absorption peak spectral line refers to the spectral curve that has been preprocessed, such as baseline correction, to eliminate interference from non-target signals, such as background noise and instrument drift. The absorption peak intensity vector is a one-dimensional vector formed by arranging the intensity values ​​of each characteristic absorption peak in the pure spectral line in wavelength order, which is used to quantify the absorption capacity of the target gas at different characteristic wavelengths.

[0058] In a specific implementation, a polynomial fitting method can be used to fit the low absorption region at both ends of the spectrum line by the least square method to generate a baseline function, and then subtract the baseline function from the original spectrum line to obtain a pure absorption peak spectrum line eliminating background interference. A peak height threshold (such as 0.2 times the maximum peak height) and a minimum peak width (such as 0.05 μm) are set to scan the spectrum line in a sliding window, identify the local maximum value points, record the wavelength position and intensity value of each peak value, and construct an absorption peak intensity vector. For example, in natural gas pipeline leakage monitoring, the absorption peak intensity of 0.85 at 3.3 μm is extracted from the corrected spectrum line, which corresponds to a methane concentration of about 5% LE. The absorption peak intensity vector is substituted into the Lambert-Beer model to calculate the initial concentration value of a single gas, wherein the Lambert-Beer model can be represented by the following formula: ; wherein, represents the initial concentration value of a single gas, represents the gas absorption coefficient, represents the optical path, represents the incident light intensity, represents the transmitted light intensity. Further, the single gas concentration value can be calculated by establishing an interference correction matrix and then using the interference correction matrix in combination with the following formula: ; wherein, represents the corrected initial concentration value of a single gas, represents the initial concentration value of a single gas, represents the interference correction matrix.

[0059] Further, the single gas concentration value is used to perform gas safety monitoring on the construction area to obtain a second safety monitoring result, which can determine whether there is a risk of leakage, exceeding the standard, etc. in the construction area (generally the underground pipeline part of the construction area).

[0060] In a specific implementation, the single gas concentration value measured in the construction area can be compared with a pre-set gas safety threshold (such as the lower limit of methane explosion and the critical value of carbon monoxide poisoning), and if the concentration exceeds the threshold, a safety warning is issued, and if it does not exceed, it is determined to be safe, thereby obtaining the second safety monitoring result. For example, when the methane concentration in the tunnel construction is detected to be 1.5% LEL (higher than the 1% warning value), a conclusion of gas insecurity is directly output and an alarm is triggered.

[0061] S4, track the hot spot trajectory of the construction equipment in the construction area by using the infrared sensing data, obtain a hot spot evolution trajectory, construct a thermal map of the construction equipment based on the hot spot evolution trajectory, and perform equipment overheating safety monitoring on the construction equipment based on the thermal map to obtain a third safety monitoring result.

[0062] By tracking the hot spot trajectory of the construction equipment in the construction area by using the infrared sensing data, the embodiment of the application can capture the infrared radiation signal of the abnormal temperature area on the surface of the equipment, record the position change of the hot spot in real time and generate a dynamic trajectory, so as to identify the overheating fault hidden danger of the equipment.

[0063] The hot spot evolution trajectory refers to a trajectory record of the dynamic evolution process of the position movement, range expansion or contraction, and temperature change of the abnormal temperature area (hot spot) on the surface of the construction equipment in the time dimension, which can intuitively reflect the development path of the equipment thermal fault.

[0064] As an embodiment of the application, tracking the hot spot trajectory of the construction equipment in the construction area by using the infrared sensing data to obtain a hot spot evolution trajectory includes: constructing a temperature matrix of the construction area by using the infrared sensing data; identifying hot spots of the construction equipment in the construction area based on the temperature matrix; performing multi-frame trajectory association on the hot spots to obtain a hot spot evolution trajectory.

[0065] The temperature matrix refers to a two-dimensional numerical matrix arranged according to the spatial positions of the pixels of the thermal imaging data of the construction area collected by the infrared sensor, and the hot spot refers to a pixel point or a connected pixel area in the temperature matrix whose temperature exceeds a preset threshold (such as the average temperature of the area + 20%) and represents an abnormally high temperature part on the surface of the construction equipment.

[0066] In a specific implementation, the thermal imaging data of the construction area collected by the infrared sensor can be digitally processed, and the temperature value corresponding to each pixel can be arranged into a two-dimensional matrix according to the spatial position of the pixel. A temperature threshold (such as 20% higher than the average temperature of the area) is set, and then each element in the temperature matrix is traversed, and the pixel points exceeding the threshold are marked as hot spots. A hot spot database on the time sequence is established, the position similarity between adjacent frames of hot spots is calculated by using the Hungarian algorithm, the hot spots with the highest similarity are associated as the same target, the coordinate change of each hot spot in multiple frames is recorded, and a continuous motion trajectory is generated.

[0067] Further, the embodiment of the present application can convert the dynamic trajectory of temperature abnormal points of each part of the equipment into a visual temperature distribution cloud map and intuitively present the overall thermal state of the equipment by constructing the thermal map of the construction equipment based on the thermal spot evolution trajectory.

[0068] The thermal map refers to converting the temperature distribution of the surface of the construction equipment into a visual color spectrum based on thermal spot evolution trajectory data, intuitively representing the temperature of each part of the equipment and the range of abnormal heat zones through different colors (such as red, yellow, and blue), and presenting an intuitive image of the overall thermal state of the equipment and potential fault risks.

[0069] In specific implementation, the position and temperature value of the temperature abnormal point at each time in the thermal spot evolution trajectory can be mapped to the corresponding part of the three-dimensional model of the construction equipment, and a continuous temperature distribution cloud map can be generated through an interpolation algorithm to form a visual thermal map.

[0070] Further, the embodiment of the present application can obtain a third safety monitoring result by monitoring the overheating safety of the construction equipment based on the thermal map.

[0071] The third safety monitoring result refers to the final safety determination of the overheating risk of the construction equipment through deep analysis (such as temperature evolution trend, heat zone diffusion speed, and equipment operation condition coupling analysis) of the thermal map data.

[0072] In specific implementation, the temperature distribution of each region of the equipment in the thermal map can be compared with the preset safety temperature threshold, the range, temperature gradient, and evolution trend of the high-temperature region can be analyzed, the overheating risk level can be determined in combination with the equipment operation condition, and disposal suggestions can be output to obtain the third safety monitoring result.

[0073] S5, based on the first safety monitoring result, the second safety monitoring result and the third safety monitoring result, constructing a safety monitoring report of the building construction area.

[0074] The safety monitoring report of the building construction area is constructed based on the first safety monitoring result, the second safety monitoring result and the third safety monitoring result, so that the dispersed risk information such as steel structure safety, gas safety and equipment overheating in the building construction area can be integrated and summarized, and the overall safety situation is presented in a structured report form, thereby providing a comprehensive and accurate decision basis for the user. For example, in subway tunnel construction, the report summarizes the methane concentration exceeding standard early warning, shield machine motor overheating risk and the like, and intuitively displays the safety hidden danger distribution of the construction area, so that the rectification scheme can be quickly formulated.

[0075] In the implementation, the first, second and third safety monitoring results (such as gas concentration abnormal points, equipment hot spot tracks, thermal map risk levels) can be integrated according to time sequence and spatial position to form a structured report containing risk type, position, level and disposal suggestion. For example, the methane concentration exceeding standard record in subway construction, the shield machine motor overheating thermal map and disposal instruction are summarized to generate a visual report with risk thermal distribution.

[0076] Embodiment 2 As shown in Figure 2 is a functional module diagram of a building construction safety monitoring system based on a sensor according to the present application.

[0077] The building construction safety monitoring system based on a sensor 200 according to the present application can be installed in an electronic device. According to the implemented functions, the building construction safety monitoring system based on a sensor can include a data acquisition module 201, a first monitoring module 202, a second monitoring module 203, a third monitoring module 204 and a safety monitoring report module 205. The modules according to the present application can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can complete a fixed function, and are stored in the memory of the electronic device.

[0078] In the embodiments of the present application, the functions of each module / unit are as follows: The data acquisition module 201 is configured to acquire multi-sensor data of a building construction area, wherein the multi-sensor data includes structural sensor data, infrared sensor data and gas sensor data, and the structural sensor data includes steel structure vibration data, steel structure inclination data and steel structure stress data. The first monitoring module 202 is configured to convert the structural sensor data into a joint embedding vector, perform structural stress deployment on the building construction area by using the joint embedding vector, obtain a dynamic twin scene, draw a structural deformation curve of the building construction area by using the dynamic twin scene, and perform steel structure safety monitoring on the building construction area by using the structural deformation curve to obtain a first safety monitoring result. The second monitoring module 203 is configured to perform mixed gas decoupling on the construction area based on the gas sensing data to obtain single gas characteristic spectral lines, analyze single gas concentration values of the multi-gas area by using the single gas characteristic spectral lines, perform gas safety monitoring on the construction area by using the single gas concentration values, and obtain a second safety monitoring result. The third monitoring module 204 is configured to perform hot spot trajectory tracking on the construction equipment of the construction area by using the infrared sensing data to obtain a hot spot evolution trajectory, construct a heat map of the construction equipment based on the hot spot evolution trajectory, perform equipment overheating safety monitoring on the construction equipment based on the heat map, and obtain a third safety monitoring result. The safety monitoring report module 205 is configured to construct a safety monitoring report of the construction area based on the first safety monitoring result, the second safety monitoring result, and the third safety monitoring result.

[0079] In detail, the modules in the construction safety monitoring system 200 in the embodiment of the present application are used in the same way as the technical means of the construction safety monitoring method in the embodiment of the present application, and can produce the same technical effects, which will not be described here. Figure 1 In detail, the modules in the construction safety monitoring system 200 in the embodiment of the present application are used in the same way as the technical means of the construction safety monitoring method in the embodiment of the present application, and can produce the same technical effects, which will not be described here.

[0080] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A sensor-based construction safety monitoring method, characterized by, The method comprises: Collecting multi-sensor data of a building construction area, wherein the multi-sensor data comprises structural sensing data, infrared sensing data and gas sensing data, and the structural sensing data comprises steel structure vibration data, steel structure inclination data and steel structure stress data; Converting the structural sensing data into a joint embedding vector, using the joint embedding vector to perform structural stress deployment on the building construction area to obtain a dynamic twin scene, using the dynamic twin scene to draw a structural deformation curve of the building construction area, using the structural deformation curve to perform steel structure safety monitoring on the building construction area to obtain a first safety monitoring result; Based on the gas sensing data, decoupling mixed gas in the building construction area to obtain single gas characteristic spectral lines, using the single gas characteristic spectral lines to analyze the single gas concentration values of the multi-gas area, and using the single gas concentration values to perform gas safety monitoring on the building construction area to obtain a second safety monitoring result; Using the infrared sensing data to track the thermal spot trajectory of the construction equipment in the building construction area to obtain a thermal spot evolution trajectory, based on the thermal spot evolution trajectory, constructing a thermal map of the construction equipment, based on the thermal map, performing equipment overheating safety monitoring on the construction equipment to obtain a third safety monitoring result; Based on the first safety monitoring result, the second safety monitoring result and the third safety monitoring result, constructing a safety monitoring report of the building construction area.

2. The sensor-based construction safety monitoring method of claim 1, wherein, Converting the structural sensing data into a joint embedding vector comprises: Extracting three-dimensional features of the structural sensing data to obtain vibration vectors, inclination vectors and stress vectors; Performing wavelet packet transform processing on the vibration vectors to obtain wavelet transform vectors; Performing quaternion differential processing on the inclination vectors to obtain four-dimensional angle vectors; Performing principal stress analysis processing on the stress vectors to obtain joint stress vectors; Performing cross-modal mapping on the wavelet transform vectors, the four-dimensional angle vectors and the joint stress vectors to obtain preliminary fusion feature vectors; Performing spatial semantic constraint on the preliminary fusion feature vectors to obtain joint embedding vectors.

3. The sensor-based construction safety monitoring method of claim 1, wherein, Using the joint embedding vector to perform structural stress deployment on the building construction area to obtain a dynamic twin scene comprises: Performing mechanical parameter decoupling on the joint embedding vector to obtain a structured mechanical parameter set; Using the structured mechanical parameter set to analyze the transient mechanical field of the building construction area; According to the transient mechanical field, constructing a stress-geometry coupling model of the building construction area; Performing visual rendering on the stress-geometry coupling model to obtain a dynamic twin scene.

4. The sensor-based construction safety monitoring method of claim 1, wherein, Using the dynamic twin scene to draw a structural deformation curve of the building construction area comprises: Performing stress gradient clustering processing on the dynamic twin scene to identify a high stress node group in the dynamic twin scene; Constructing a displacement field matrix of stress nodes in the high stress node group; Based on the displacement field matrix, calculating the node equivalent deformation variable in the dynamic twin scene; Based on the node equivalent deformation variable, drawing a structural deformation curve of the building construction area.

5. The sensor-based construction safety monitoring method of claim 1, wherein, The structure deformation curve is used for steel structure safety monitoring of the building construction area, and a first safety monitoring result is obtained, including: calculating the deformation rate of the structure deformation curve; based on the deformation rate, identifying the steel structure mutation point in the building construction area to obtain an abnormal point set; according to the abnormal point set, identifying the steel structure abnormal mode of the building construction area; based on the steel structure abnormal mode, the steel structure safety monitoring of the building construction area is carried out, and the first safety monitoring result is obtained.

6. A sensor-based construction safety monitoring method as claimed in claim 5, characterized in that, The calculation of the deformation rate of the structure deformation curve includes: sliding window processing is performed on the structure deformation curve to obtain a sampling curve; the deformation rate of the sampling curve is calculated by using the following formula: ； wherein, denotes a deformation rate of a sampling curve, denotes an i-th sliding window sampling timestamp, n denotes a sliding window radius, and k denotes a relative position index within the sliding window, denotes a boundary position is extended backward by n time steps, denotes a boundary position is extended forward by n time steps, denotes a Gaussian weight of k, denotes a sampling interval, denotes 1 a deformation variable at time t, denotes a deformation variable at time t.

7. The sensor-based construction safety monitoring method of claim 1, wherein, based on the gas sensing data, the mixed gas decoupling of the building construction area is carried out, and a single gas characteristic spectrum line is obtained, including: the gas sensing data is standardized to obtain standardized gas data; using the standardized gas data, a mixed gas feature matrix of the building construction area is constructed; using the mixed gas feature matrix, the absorption peak intensity and wavelength of the mixed gas in the building construction area are identified; based on the absorption peak intensity and the wavelength, the spectral signal intensity of the mixed gas is calculated by using the following formula: ； wherein, spectral signal intensity, N represents the number of mixed gases, represents the absorption peak intensity of the jth gas in the mixed gas, wavelength of the mixed gas, represents the central wavelength of j, represents the standard deviation of j; based on the spectral signal intensity, the mixed gas decoupling of the building construction area is carried out, and a single gas characteristic spectrum line is obtained.

8. The sensor-based construction safety monitoring method of claim 1, wherein, Using the single gas characteristic spectrum line, the single gas concentration value of the multi-gas area is analyzed, including: baseline correction is performed on the single gas characteristic spectrum line to obtain a pure absorption peak spectrum line; peak value determination is performed on the pure absorption peak spectrum line to obtain an absorption peak intensity vector at a specific wavelength in the pure absorption peak spectrum line; based on the absorption peak intensity vector, the initial concentration value of the single gas in the multi-gas area is identified by using the Lambert-Beer calibration model; cross interference correction is performed on the initial concentration value of the single gas to obtain the single gas concentration value.

9. The sensor-based construction safety monitoring method of claim 1, wherein, Using the infrared sensing data, the hot spot trajectory tracking of the construction equipment in the building construction area is performed to obtain a hot spot evolution trajectory, including: using the infrared sensing data, a temperature matrix of the building construction area is constructed; based on the temperature matrix, the hot spot of the construction equipment in the building construction area is identified; multi-frame trajectory correlation is performed on the hot spot to obtain a hot spot evolution trajectory.

10. A sensor-based construction safety monitoring system, characterized by, The system includes: a data acquisition module for acquiring multi-sensor data of a building construction area, wherein the multi-sensor data includes structure sensing data, infrared sensing data and gas sensing data, and the structure sensing data includes steel structure vibration data, steel structure inclination data and steel structure stress data; a first monitoring module for converting the structure sensing data into a joint embedding vector, using the joint embedding vector for structure stress deployment of the building construction area, obtaining a dynamic twin scene, using the dynamic twin scene to draw a structure deformation curve of the building construction area, using the structure deformation curve for steel structure safety monitoring of the building construction area, and obtaining a first safety monitoring result; The second monitoring module is configured to perform mixed gas decoupling on the construction area based on the gas sensing data to obtain single gas characteristic spectral lines, analyze single gas concentration values of the multi-gas area by using the single gas characteristic spectral lines, perform gas safety monitoring on the construction area by using the single gas concentration values, and obtain a second safety monitoring result. The third monitoring module is configured to perform hot spot trajectory tracking on the construction equipment of the construction area by using the infrared sensing data to obtain a hot spot evolution trajectory, construct a heat map of the construction equipment based on the hot spot evolution trajectory, perform equipment overheating safety monitoring on the construction equipment based on the heat map, and obtain a third safety monitoring result. The safety monitoring report module is configured to construct a safety monitoring report of the construction area based on the first safety monitoring result, the second safety monitoring result, and the third safety monitoring result.

Images