Steel roof supporting structure safety monitoring method and system based on visual identification

Abstract

The invention relates to a steel roof supporting structure safety monitoring method and system based on visual identification, and belongs to the field of constructional engineering, and the method comprises the steps: building a finite element model of a steel roof supporting structure; according to nodes of the finite element model, monitoring points are selected on the steel roof supporting structure; the real-time displacement of a monitoring point is obtained through visual recognition, and the method comprises the steps that a marker dynamically adaptive to the environment is arranged on the monitoring point; collecting and identifying a plurality of marker images through a calibration camera, and obtaining real-time displacements of a plurality of monitoring points; adjusting a real-time displacement alarm threshold value of the monitoring point according to the real-time working condition When the real-time displacement of any monitoring point exceeds the corresponding alarm threshold value, an alarm signal is sent out. The monitoring efficiency is improved, the method is suitable for various field illumination environments, the real-time displacement of a monitoring point can be obtained more accurately and timely, corresponding alarm threshold values can be set after the stress data of the steel roof supporting structure changes under different working conditions, and more effective safety monitoring is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of building engineering, and in particular to a method and system for safety monitoring of a steel roof support structure based on visual recognition. Background Art

[0002] With the development of modern architecture, large-span and high-space building forms are gradually increasing, and steel roof support structure systems are also widely used in buildings such as airports and exhibition centers. However, since the support systems of these buildings usually have complex structures, large loads, and high construction difficulties, safety monitoring faces severe challenges. At present, support structure monitoring mostly relies on pressure and displacement sensors. Due to the number and quality of physical sensors, there are fewer data collection points, and the monitoring accuracy and timeliness are limited. It is difficult to fully grasp the stress and deformation of the support structure in real time. In addition, the monitoring thresholds of traditional methods mostly rely on experience settings, which is difficult to accurately reflect the actual needs of each specific project. Therefore, there is an urgent need for a more efficient, more accurate and better real-time monitoring solution to improve the construction safety of steel roof support structures. Summary of the invention

[0003] In order to solve the above problems and improve the safety monitoring of steel roof supporting structures, the inventors have made the present invention to provide a method and system for safety monitoring of steel roof supporting structures based on visual recognition.

[0004] In a first aspect, the present invention provides a method for safety monitoring of a steel roof support structure based on visual recognition, comprising the following steps: establishing a finite element model of the steel roof support structure; selecting monitoring points on the steel roof support structure according to the nodes of the finite element model; obtaining the real-time displacement of the monitoring points through visual recognition, including: setting markers that dynamically adapt to the environment at the monitoring points; acquiring the real-time displacement of multiple monitoring points by collecting and identifying multiple marker images with a calibrated camera; adjusting the real-time displacement alarm threshold of the monitoring points according to the real-time working conditions; and issuing an alarm signal when the real-time displacement of any monitoring point exceeds the corresponding alarm threshold.

[0005] Specifically, establishing a finite element model of a steel roof supporting structure includes the following steps: constructing a finite element model of the steel roof supporting structure according to the specific structure of the Y-shaped column of the steel roof supporting structure, the specific structure including bottom steel columns, eight-variable and four-variable section columns, cast steel nodes, and bifurcated columns; simulating the stress state of the Y-shaped column through simulation and multi-condition stress analysis, and determining the stress performance of the joint axis and connection nodes of the Y-shaped column; obtaining the displacement characteristics and buckling loads of the nodes of the steel roof supporting structure according to the stress performance of the joint axis and connection nodes of the Y-shaped column; setting the real-time displacement alarm threshold of the monitoring point during the construction process according to the displacement characteristics and buckling loads of the nodes of the steel roof supporting structure.

[0006] Specifically, the marker that dynamically adapts to the environment emits light according to a color sequence or frequency coding, and adjusts the marker's light intensity according to the intensity of the on-site light.

[0007] Specifically, the real-time displacement of multiple monitoring points is obtained by collecting and identifying multiple marker images through a calibrated camera, including the following steps: collecting an image including multiple markers through a calibrated camera; visually identifying the markers in the image to obtain the image pixel coordinates of the multiple markers; obtaining the actual physical coordinates of the multiple markers based on the conversion relationship between the image pixel coordinates and the actual physical coordinates; and determining the real-time displacement of the multiple monitoring points based on the changes in the actual physical coordinates of the multiple markers.

[0008] Specifically, the safety monitoring method of the steel roof supporting structure based on visual recognition also includes the following steps: correcting the real-time displacement of the corresponding monitoring point according to the multi-source data of the monitoring point; the multi-source data includes: obtaining the real-time displacement of the monitoring point through visual recognition, and obtaining the real-time displacement of the monitoring point through different sensors respectively.

[0009] Specifically, the safety monitoring method of the steel roof supporting structure based on visual recognition also includes the following steps: according to the real-time displacement of the monitoring point, the parameters of the finite element model are corrected; the finite element model after the parameter correction is subjected to buckling analysis to obtain the buckling failure displacement of the monitoring point; the buckling failure displacement of the monitoring point is compared with the real-time displacement alarm threshold, if the real-time displacement alarm threshold of the monitoring point is greater than the buckling failure displacement of the corresponding monitoring point, the real-time displacement alarm threshold of the monitoring point is adjusted to be less than the buckling failure displacement of the corresponding monitoring point.

[0010] Specifically, the parameters of the modified finite element model include: elastic modulus of the steel roof support structure, node connection stiffness and support position of the Y-column.

[0011] Specifically, the safety monitoring method of the steel roof supporting structure based on visual recognition also includes the following steps: establishing a digital twin model of the steel roof supporting structure to obtain the real-time displacement of the monitoring points at the construction site; using the digital twin model to simulate the construction plan, and predicting the monitoring points and construction stages with construction hazards by comparing the real-time displacement alarm threshold of the monitoring points with the real-time displacement of the monitoring points at the construction site.

[0012] Specifically, the method for safety monitoring of steel roof supporting structures based on visual recognition also includes the following steps: optimizing the construction plan or the real-time displacement alarm threshold of the monitoring point according to the monitoring points and construction stages where construction hazards exist.

[0013] In a second aspect, the present invention provides a steel roof support structure safety monitoring system based on visual recognition, comprising: The monitoring module is used to establish a finite element model of the steel roof support structure; select monitoring points on the steel roof support structure according to the nodes of the finite element model; obtain the real-time displacement of the monitoring points through visual recognition, including: setting markers that dynamically adapt to the environment at the monitoring points; acquiring the real-time displacement of multiple monitoring points by collecting and identifying multiple marker images through a calibrated camera; and adjusting the real-time displacement alarm threshold of the monitoring points according to the real-time working conditions.

[0014] The alarm module is used to send out an alarm signal when the real-time displacement of any monitoring point exceeds the corresponding alarm threshold.

[0015] The beneficial effects of the above technical solution provided by the present invention include at least: (1) The present invention can perform safety monitoring on the steel roof supporting structure. By setting markers that dynamically adapt to the environment at the monitoring points and collecting and identifying multiple marker images through a calibrated camera, the real-time displacement of multiple monitoring points can be obtained. This can reduce the number of physical sensors used, save monitoring costs, and facilitate a large increase in the number of monitoring points. More finite element model nodes can be monitored, which is beneficial for adjusting the position and number of monitoring points under different working conditions, improving monitoring efficiency, and being able to monitor spaces where it is difficult to install sensors.

[0016] (2) By adopting markers that dynamically adapt to the environment, the present invention is applicable to a variety of on-site lighting environments, making it easier for the camera to collect and identify marker images, which is conducive to more accurate and timely acquisition of the real-time displacement of the monitoring point and avoiding the impact of the difficult identification of the markers on monitoring.

[0017] (3) The present invention adjusts the real-time displacement alarm threshold of the monitoring point according to the real-time working conditions. After the stress data of the steel roof support structure changes under different working conditions, the corresponding alarm threshold can be set to achieve more effective safety monitoring.

[0018] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0019] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1It is a flow chart of a method for safety monitoring of a steel roof support structure based on visual recognition in an embodiment of the present invention; Figure 2 A diagram of monitoring point markers identified by computer vision in an embodiment of the present invention; Figure 3 This is a flowchart of computer vision real-time monitoring alarm in an embodiment of the present invention; Figure 4 This is a digital twin model diagram of the steel roof support structure in an embodiment of the present invention. DETAILED DESCRIPTION

[0021] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0022] In order to solve the problems existing in the prior art, the embodiment of the present invention provides a method and system for safety monitoring of steel roof support structure based on visual recognition, which performs safety monitoring on the construction process of steel roof support structure, and realizes high-precision real-time monitoring, alarm and dynamic adjustment of steel roof support structure by combining finite element model and computer vision technology. This implementation method not only ensures the safety of the construction process, but also provides convenient remote monitoring means and rich data support for management personnel, meeting the high requirements of modern construction industry for safety monitoring.

[0023] The embodiment of the present invention provides a method for safety monitoring of a steel roof support structure based on visual recognition, the process of which is as follows: Figure 1 As shown, the following steps are included: Step S1: Establish a finite element model of the steel roof support structure; select monitoring points on the steel roof support structure according to the nodes of the finite element model; obtain the real-time displacement of the monitoring points through visual recognition, including: setting markers that are dynamically adaptive to the environment at the monitoring points; collect and identify multiple marker images through a calibrated camera to obtain the real-time displacement of multiple monitoring points; adjust the real-time displacement alarm threshold of the monitoring points according to the real-time working conditions.

[0024] Traditional monitoring methods often rely on dispersed sensors to collect data. This method has problems with data collection delays and transmission lags, and can only obtain a small amount of data at a fixed location. The present invention uses a computer vision system to obtain image data of the supporting structure in real time. The computer vision algorithm can quickly identify and analyze the real-time position changes of the markers, complete data collection and displacement calculations in a timely manner, and ensure the real-time nature of the monitoring data. At the same time, the alarm mechanism in the system can immediately respond to displacement anomalies. Once the displacement of a key node is detected to exceed the threshold, the system automatically issues an alarm to prompt construction personnel to stop work and inspect. Through the rapid response mechanism, the present invention can provide timely warnings before an accident occurs, greatly reducing the probability of an accident.

[0025] The present invention avoids the need to install a large number of sensors by setting monitoring points and arranging markers at the key nodes of the supporting structure, thereby reducing the time cost of on-site installation and debugging. At the same time, the computer vision system can monitor the displacement changes of multiple nodes at the same time, breaking the limitations of single-point monitoring of traditional methods, realizing synchronous monitoring of multiple nodes, and further improving the monitoring efficiency of the system. The finite element model is established according to the structural design requirements before the project starts, and the initial alarm threshold is determined through simulation. During the construction process, the model is corrected according to the actual displacement data, so that the alarm value is gradually optimized. Compared with traditional methods, the efficiency of the present invention is not only reflected in construction safety monitoring, but also in the real-time nature of the data and the adaptive adjustment of the model, ensuring that the most accurate early warning information is obtained in the shortest time, effectively improving the response efficiency and decision-making ability of construction safety management.

[0026] In some specific embodiments, a finite element model of a steel roof support structure is established, comprising the following steps: according to the specific structure of the Y-shaped column of the steel roof support structure, a finite element model of the steel roof support structure is constructed, wherein the specific structure includes a bottom steel column, an eight-variable four-variable cross-section column, a cast steel node, and a bifurcated column. The established finite element model of the steel roof support structure should fully consider the connection mode and geometric structure characteristics of the support structure to ensure that the finite element simulation results can reflect the actual stress conditions. Through simulation and multi-condition stress analysis, the stress state of the Y-shaped column is simulated, the stress and deformation conditions that may occur in the support system during construction are simulated, the connection mode and mechanical properties of each node are simulated, especially the stress state of the large joint axis, and the stress performance of the Y-shaped column joint axis and the connection node is determined. According to the stress performance of the Y-shaped column joint axis and the connection node, the displacement characteristics and buckling load of the steel roof support structure node are obtained. According to the displacement characteristics and buckling load of the steel roof support structure node, the real-time displacement alarm threshold of the monitoring point during construction is set. Specifically, the actual geometric dimensions, materials and connection relationships of each component are input into the model so that stress analysis and buckling simulation can be performed under different working conditions (such as maximum load, wind load, etc.), thereby obtaining the displacement variation range and buckling load value of key nodes. By analyzing the stress, deformation and displacement of the structure under different working conditions, the alarm thresholds of key nodes during the construction process are determined.

[0027] The model determines the vertical displacement of the supporting structure at the key nodes through buckling analysis under various construction conditions, and obtains the displacement value of the key node when the supporting structure is buckled and damaged, so as to determine the alarm threshold of construction monitoring.

[0028] In some specific embodiments, the marker that dynamically adapts to the environment emits light according to a color sequence or frequency coding, and adjusts the marker's light intensity according to the intensity of the on-site light.

[0029] Install cameras at key nodes (such as the joint nodes of the Y-shaped columns) of the steel roof support structure, calibrate the camera of the monitoring system, and select monitoring points to paste markers for visual identification. Figure 2 As shown, the marker has a reflective or unique pattern, which is convenient for the computer vision system to identify and locate under different lighting conditions. Figure 2 The markers shown, Figure 2 The pattern shown is a unique pattern for easy image recognition. Figure 2The pattern shown does not represent any meaning. The image data of the supporting structure is obtained by installing a camera at the construction site, and the camera calibration is combined with image recognition to achieve real-time displacement monitoring of the steel roof supporting structure. The target is tracked by the computer vision algorithm to obtain the displacement data of the key nodes in real time. The markers are pasted at each key node, and the real-time position of the markers is identified by the computer vision system to monitor the displacement changes at the Y-column nodes.

[0030] Furthermore, dynamic adaptive marker technology is introduced: using actively luminous LED markers, combined with color or frequency coding technology, to maintain high recognizability in low-light or highly reflective environments. For example, the markers emit light in red, blue, green, and blue as a unit sequence, and different color unit sequences are set to achieve accurate distinction between markers at different monitoring points, which helps the calibration camera to simultaneously capture images including multiple markers, and identify multiple markers at the same time, thereby improving monitoring efficiency. In order to improve monitoring efficiency and for each calibration camera to simultaneously capture and identify multiple markers, the frequency of the marker's light emission or other signal emission is encoded, and different frequency codes are set for markers at different monitoring points, so that multiple markers can be identified at the same time.

[0031] Furthermore, the marker obtains the on-site light intensity through a photoelectric sensor or a photoresistor, or obtains the on-site light parameters by receiving the feedback signal of the marker image recognition system, thereby realizing the environmental perception function and determining the on-site light conditions. The corresponding relationship between the on-site light conditions and the luminous brightness of the marker is set. Then, according to the corresponding relationship between the on-site light conditions and the luminous brightness, the luminous brightness is automatically adjusted to ensure the accuracy and robustness of monitoring under different lighting conditions around the clock.

[0032] In some specific embodiments, a plurality of marker images are captured and identified by a calibrated camera to obtain real-time displacements of a plurality of monitoring points, comprising the following steps: capturing an image including a plurality of markers by a calibrated camera; visually identifying the markers in the image to obtain image pixel coordinates of the plurality of markers; obtaining the actual physical coordinates of the plurality of markers based on a conversion relationship between the image pixel coordinates and the actual physical coordinates; and determining the real-time displacements of the plurality of monitoring points based on changes in the actual physical coordinates of the plurality of markers.

[0033] Computer vision algorithms are used to track the position changes of markers and calculate the real-time displacement of each monitoring point during the construction process. The above-mentioned computer vision algorithms include target recognition, template matching, normalized correlation coefficient algorithm and camera internal and external parameter solution, which can accurately track markers under different environmental conditions to ensure that the monitoring system accurately tracks and records the displacement of each key node. Through automatically collected data, the displacement changes of key nodes are displayed in real time, and the structural status during the construction process is recorded.

[0034] Taking the template matching algorithm as an example, the identification of the labeled image is explained. During the real-time displacement monitoring process, the template matching algorithm and the normalized correlation coefficient method are used to track the target of the marker to ensure the real-time positioning of the monitoring point within the field of view. When the monitoring point deviates from the set trajectory, the system issues a position offset alarm. The template matching method is used for target tracking. First, the marker is pasted on the measuring point. In the first frame image, the area with more obvious features (that is, the area where the measuring point is located) is selected as the predefined template. Then, the predefined template is moved pixel by pixel on the region of interest (ROI) to search for the target area. The matching degree between the template and the area is calculated by the normalized correlation coefficient.

[0035] Camera calibration uses the proportional coefficient method or other appropriate image calibration methods to convert image pixel coordinates to actual physical coordinates to ensure the accuracy of real-time displacement monitoring and meet the needs of precision monitoring of construction.

[0036] Taking the proportional coefficient method as an example, the conversion between image pixel coordinates and actual physical coordinates is explained. According to the specific size and pixel distance of the Y-column support structure, the proportional coefficient of each monitoring point is calculated, and the position of the monitoring point in the image is corrected to obtain accurate actual displacement data. Calculate the proportional coefficient s at the monitoring point, s = d / p, where p is the image pixel distance between the two points and d is the actual physical distance between the two points. Multiply the calculated proportional coefficient s by the actual physical distance to obtain the image pixel distance, and then determine the image pixel coordinates based on the reference point coordinates in the coordinate system. The actual physical distance is calculated by two or more actual physical coordinates using the Euclidean distance formula.

[0037] The flowchart of the security monitoring implementation example using visual recognition is as follows Figure 3 As shown, the measuring point (monitoring point) is selected, the proportional coefficient between the image pixel coordinates and the actual physical coordinates is determined, the real-time construction image is obtained, the real-time displacement is calculated by identifying the position of the monitoring point, and the real-time displacement of the monitoring point is compared with the alarm value (alarm threshold). If the real-time displacement of the monitoring point is greater than the alarm value, it indicates that the real-time displacement exceeds the safety limit, an over-limit alarm signal is issued, and the work is stopped for rectification. The above method is used to continuously monitor the construction process until the construction is completed.

[0038] In some specific embodiments, the above-mentioned method for safety monitoring of steel roof support structure based on visual recognition further includes the following steps: according to the multi-source data of the monitoring point, correcting the real-time displacement of the corresponding monitoring point; the multi-source data includes: obtaining the real-time displacement of the monitoring point through visual recognition, and obtaining the real-time displacement of the monitoring point through different sensors respectively. The different sensors can be sensors of the same type, such as multiple strain sensors, or multiple sensors of different types, such as multiple sensors such as strain sensors, acceleration sensors and laser rangefinders.

[0039] For example, to further improve the accuracy and robustness of monitoring, the system will integrate multimodal sensor data, and integrate the real-time displacements obtained by visual recognition, strain sensors, acceleration sensors and laser ranging data by building an intelligent fusion platform. The intelligent fusion platform uses the Kalman filter algorithm to preprocess multi-source data to achieve dynamic optimization of monitoring results. At the same time, a self-calibration mechanism is introduced. During the construction process, the visual monitoring error is automatically corrected through cross-validation of different sensor data, and the monitoring logic is adjusted in real time through intelligent identification of sensor failures. For example, the real-time displacement after data fusion is obtained by weighted summation of the real-time displacements measured by multiple sensors. When any of the sensors continuously sends out abnormal values, the weight of the corresponding sensor data is reset to zero. At the same time, the weight of sensor data is dynamically adjusted, such as reducing the weight of sensor data with large errors and increasing the weight of sensor data with small errors, so as to achieve a small error in real-time displacement after data fusion. Through the above data fusion design, the comprehensive performance of the system and the reliability of real-time monitoring will be significantly improved.

[0040] By using the above-mentioned markers with higher recognition and the above-mentioned high-precision image recognition algorithm, it is helpful to stably track the position of the markers under different lighting conditions, viewing angle offset, etc., and accurately obtain the real-time displacement data of each monitoring point. Compared with installing sensors, markers are easier to set up and the range of monitoring points is large, so they have higher monitoring flexibility and adaptability.

[0041] In some specific embodiments, the above-mentioned method for monitoring the safety of the steel roof support structure based on visual recognition further includes the following steps: according to the real-time displacement of the monitoring point, the parameters of the finite element model are corrected; the finite element model after the parameter correction is subjected to buckling analysis to obtain the buckling failure displacement of the monitoring point; the buckling failure displacement of the monitoring point is compared with the real-time displacement alarm threshold, and if the real-time displacement alarm threshold of the monitoring point is greater than the buckling failure displacement of the corresponding monitoring point, the real-time displacement alarm threshold of the monitoring point is adjusted to be less than the buckling failure displacement of the corresponding monitoring point. The alarm threshold after the model parameter correction is recalculated to ensure the accuracy and applicability of the system during the alarm.

[0042] Furthermore, the parameters of the modified finite element model include factors such as the elastic modulus of the steel roof support structure, the node connection stiffness and the support position of the Y-column. Modifying the above parameters can make the finite element model closer to the stress state under actual working conditions.

[0043] During the construction process, the material properties, node connection stiffness, dimensional parameters of the supporting components, and other characteristics of the finite element model are corrected based on the real-time displacement data collected by the computer vision system. That is, the monitoring data during the construction period (such as actual displacement and deviation data) is input into the finite element model to correct the material parameters, structural stiffness, node connection characteristics, etc. of the model. The finite element model with parameter correction is re-analyzed for buckling, and the rationality of the current alarm threshold is verified. The corrected model can more accurately reflect the actual stress conditions, provide data support for the adaptive adjustment of the model, and can be dynamically adjusted as the construction conditions change, making the alarm threshold setting more accurate and ensuring that the alarm threshold is set reasonably, which helps to reduce the possibility of false alarms and missed alarms and improve the reliability of safety monitoring of the supporting structure.

[0044] Based on the alarm thresholds obtained by finite element simulation, a real-time monitoring alarm mechanism is set at each key node. When the computer vision system detects that the displacement value of any key node exceeds the alarm threshold, the system automatically sends an alarm signal to remind construction personnel to immediately check the status of the stress node and take emergency measures.

[0045] The safety monitoring system automatically records the real-time displacement data of each monitoring point and other monitoring data, alarm information and model correction parameters during the entire construction process, and stores them in the system database for subsequent analysis of the mechanical properties of the steel roof support structure throughout its life cycle. After the construction is completed, the historical monitoring data in the database will be used to analyze the displacement changes and stress characteristics of the support structure at different construction stages, providing a reference for the monitoring and maintenance of the support structure throughout its life cycle. Through trend analysis of displacement data, the characteristics of the support structure under different loads can be identified, and the monitoring model can be improved or the construction process of the support structure can be optimized accordingly. In addition, data records provide an objective basis for subsequent safety assessments and project summaries, and can be used for experience accumulation and management decisions for similar projects.

[0046] Real-time data is sent to the remote monitoring center via a wireless network to achieve remote monitoring and store alarm data for subsequent analysis and construction management review. Construction management personnel can view the stress state, displacement changes and alarm information of the steel roof support structure in real time in a control room outside the construction site or in a remote office environment. This remote monitoring function not only improves management efficiency and enables managers to grasp the progress of construction anytime and anywhere, but is also particularly important for the centralized management of large projects. For multiple projects distributed across multiple construction sites, construction management can centrally monitor through a set of systems to improve management convenience and response speed. At the same time, the automatic storage and history recording functions of remote monitoring data also provide reliable data support for subsequent project analysis, management decisions and construction quality tracking.

[0047] In some specific embodiments, the above-mentioned steel roof support structure safety monitoring method based on visual recognition also includes the following steps: establishing a digital twin model of the steel roof support structure to obtain the real-time displacement of the monitoring points at the construction site; using the digital twin model to simulate the construction plan, and by comparing the real-time displacement alarm threshold of the monitoring points with the real-time displacement of the monitoring points at the construction site, predicting the monitoring points and construction stages with construction hazards. During the construction process of a steel roof support structure, a three-dimensional screenshot retrieved from the established digital twin model is shown in FIG. Figure 4 As shown, in the digital twin model, by selecting different fields, stereo images with different data features can be generated and displayed. Figure 4 Screenshot of the digital twin model shown.

[0048] Based on the interaction of virtual and real information through the digital twin model, the information synchronization between the physical space and the virtual space is ensured. Through this interactive process, the real-time data of the physical site can be quickly updated to the virtual model (digital twin model), and the simulation results and risk assessment in the virtual model will be promptly fed back to the operators at the physical site. This two-way flow of data ensures real-time monitoring and dynamic adjustment of the lifting operation. Digital twin technology provides a comprehensive intelligent solution for the monitoring of high roof support structures through the combination of virtual and real. Based on real-time data flow, the digital twin dynamic model can reflect the key status of the construction site in real time, such as node displacement, stress distribution and load change, and simulate potential risks under complex working conditions by combining with the finite element model. At the same time, the model is trained using historical data to predict hidden dangers in the future construction stage and optimize alarm thresholds and construction strategies. Virtual reality (VR) and augmented reality (AR) technologies further enhance the interactive capabilities of digital twins. Managers can view the construction status, simulate construction plans, and respond quickly to emergencies in a virtual environment in real time. In addition, the virtual-reality bidirectional feedback mechanism realizes the interaction between on-site data and virtual models. The analysis results reversely guide construction adjustments, such as stress equalization or load distribution optimization, thereby improving construction safety and efficiency.

[0049] In some specific embodiments, the above-mentioned method for safety monitoring of steel roof support structure based on visual recognition also includes the following steps: optimizing the construction plan or the real-time displacement alarm threshold of the monitoring point according to the monitoring point and construction stage where there are construction hazards. Specifically, the system makes a comprehensive comparison of real-time data and historical data, combined with statistical analysis or machine learning algorithms, and predicts abnormal displacement or stress conditions that may occur in the future based on the historical data of the monitoring point and construction stage where there are construction hazards. When the system recognizes that the displacement change trend of certain monitoring points may exceed the safety range (for example, approaching the alarm threshold or the change rate is abnormal), an alarm is issued in advance to remind construction personnel to pay attention to potential risks.

[0050] Furthermore, the system using the digital twin model can automatically identify and manage multiple safety factors in lifting operations, including the load of the lifting equipment, the rationality of the lifting point setting, the stability of the components, and the safe operation behavior of the personnel. The system can not only monitor the potential risks such as equipment failure and personnel misoperation during the construction process, but also automatically manage the safety and provide intelligent early warning for each key operation link based on real-time monitoring data, thereby reducing the time of human intervention and improving the reliability of safety assurance.

[0051] Step S2: When the real-time displacement of any monitoring point exceeds the corresponding alarm threshold, an alarm signal is issued.

[0052] In some specific embodiments, the displacement data collected in real time is compared with the set alarm threshold. When the displacement value of any monitoring point exceeds the alarm threshold, the system automatically sends an alarm signal and sends the alarm information to the monitoring center or the manager's mobile device through a wireless network. After receiving the alarm information, the construction manager can immediately check the structural status of the relevant parts and take timely suspension or adjustment measures to eliminate potential risks. The alarm mechanism can not only issue local alarm sound and light prompts, but also has remote data transmission and alarm notification functions. All alarm events are automatically stored in the database for subsequent construction analysis and management review.

[0053] In some specific embodiments, a multi-level intelligent alarm system is proposed, and three thresholds of early warning, general alarm and emergency alarm are set. The dynamic threshold is adjusted through real-time working conditions to ensure that the monitoring results are adapted to the stress state of different construction stages. At the same time, the system supports a collaborative response mechanism, which can automatically communicate with the construction equipment to adjust the support or load distribution after the alarm signal is triggered. In addition, the system combines a visual interface and light signal prompts to realize an alarm method that integrates sound, light and vision. This optimization will significantly improve the system's sensitivity, response speed and emergency management capabilities, and realize remote safety monitoring of the steel roof support structure.

[0054] In the above method of this embodiment, the present invention can perform safety monitoring on the steel roof support structure. By setting a marker that is dynamically adaptive to the environment at the monitoring point, and collecting and identifying multiple marker images by calibrating the camera, the real-time displacement of multiple monitoring points can be obtained, which can reduce the number of physical sensors used, save monitoring costs, and facilitate a large increase in the number of monitoring points. More finite element model nodes are monitored, which is conducive to adjusting the position and number of monitoring points under different working conditions, improving monitoring efficiency, and being able to monitor spaces where it is difficult to install sensors. By adopting markers that are dynamically adaptive to the environment, the present invention is applicable to a variety of on-site lighting environments, which is convenient for calibrating the camera to collect and identify marker images, and is conducive to more accurate and timely acquisition of the real-time displacement of the monitoring point, avoiding the influence of the difficult identification of the marker on the monitoring. The present invention adjusts the real-time displacement alarm threshold of the monitoring point according to the real-time working conditions, and can set the corresponding alarm threshold after the force data of the steel roof support structure under different working conditions changes, so as to achieve more effective safety monitoring.

[0055] The above-mentioned steel roof support structure safety monitoring method is applicable to various steel roof support structure scenarios, especially the large-span structures involved in airport projects, and has high monitoring flexibility and adaptability.

[0056] Those skilled in the art can change the above sequence without departing from the protection scope of the present invention.

[0057] Another embodiment of the present invention provides a steel roof support structure safety monitoring system based on visual recognition, comprising: The monitoring module is used to establish a finite element model of the steel roof support structure; select monitoring points on the steel roof support structure according to the nodes of the finite element model; obtain the real-time displacement of the monitoring points through visual recognition, including: setting markers that are dynamically adaptive to the environment at the monitoring points; collect and identify multiple marker images through a calibrated camera to obtain the real-time displacement of multiple monitoring points; adjust the real-time displacement alarm threshold of the monitoring points according to the real-time working conditions; The alarm module is used to send out an alarm signal when the real-time displacement of any monitoring point exceeds the corresponding alarm threshold.

[0058] Regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0059] In this embodiment, the present invention can perform safety monitoring on the steel roof support structure. By setting a marker that is dynamically adaptive to the environment at the monitoring point, and collecting and identifying multiple marker images by calibrating the camera, the real-time displacement of multiple monitoring points can be obtained, which can reduce the number of physical sensors used, save monitoring costs, and facilitate a large increase in the number of monitoring points. More finite element model nodes can be monitored, which is conducive to adjusting the position and number of monitoring points under different working conditions, improving monitoring efficiency, and being able to monitor spaces where it is difficult to install sensors. By adopting markers that are dynamically adaptive to the environment, the present invention is applicable to a variety of on-site lighting environments, which is convenient for calibrating the camera to collect and identify marker images, and is conducive to more accurate and timely acquisition of the real-time displacement of the monitoring point, avoiding the influence of the difficult identification of the marker on monitoring. The present invention adjusts the real-time displacement alarm threshold of the monitoring point according to the real-time working conditions, and can set the corresponding alarm threshold after the force data of the steel roof support structure under different working conditions changes, so as to achieve more effective safety monitoring.

[0060] Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention shall still fall within the patent scope of the present invention.

Claims

1. A method for safety monitoring of steel roof support structure based on visual recognition, characterized in that: The following steps are involved: Establish a finite element model of the steel roof support structure; Selecting monitoring points on the steel roof support structure according to the nodes of the finite element model; The real-time displacement of the monitoring points is obtained through visual recognition, including: setting markers that are dynamically adaptive to the environment at the monitoring points; acquiring and identifying multiple marker images through a calibrated camera to obtain the real-time displacement of multiple monitoring points; Adjust the real-time displacement alarm threshold of the monitoring point according to the real-time working conditions; When the real-time displacement of any monitoring point exceeds the corresponding alarm threshold, an alarm signal is issued.

2. The method according to claim 1, characterized in that Establishing the finite element model of the steel roof support structure includes the following steps: According to the specific structure of the Y-shaped column of the steel roof support structure, a finite element model of the steel roof support structure is constructed, wherein the specific structure includes a bottom steel column, an eight-variable four-variable section column, a cast steel node, and a bifurcated column; Through simulation and multi-condition stress analysis, the stress state of the Y-column is simulated to determine the stress performance of the joint axis and connection node of the Y-column; According to the mechanical properties of the Y-column joint axis and connection nodes, the displacement characteristics and buckling loads of the steel roof support structure nodes are obtained; According to the displacement characteristics and buckling load of the steel roof supporting structure nodes, the real-time displacement alarm threshold of the monitoring point during the construction process is set.

3. The method according to claim 1, characterized in that The marker that dynamically adapts to the environment emits light according to a color sequence or frequency coding, and adjusts the luminous intensity of the marker according to the intensity of the on-site light.

4. The method according to claim 1, characterized in that The real-time displacement of multiple monitoring points is obtained by collecting and identifying multiple marker images through a calibrated camera, including the following steps: Acquiring an image including a plurality of landmarks by calibrating a camera; Performing visual recognition on the markers in the image to obtain image pixel coordinates of the multiple markers; Obtaining actual physical coordinates of the multiple markers according to a conversion relationship between image pixel coordinates and actual physical coordinates; The real-time displacements of the plurality of monitoring points are determined according to the changes in the actual physical coordinates of the plurality of markers.

5. The method according to claim 1, characterized in that The method for safety monitoring of the steel roof support structure based on visual recognition further comprises the following steps: According to the multi-source data of the monitoring points, the real-time displacement of the corresponding monitoring points is corrected; The multi-source data includes: obtaining the real-time displacement of the monitoring point through visual recognition, and obtaining the real-time displacement of the monitoring point through different sensors respectively.

6. The method according to claim 1, characterized in that The method for safety monitoring of the steel roof support structure based on visual recognition further comprises the following steps: According to the real-time displacement of the monitoring point, the parameters of the finite element model are modified; The buckling analysis is performed on the finite element model after parameter correction to obtain the buckling failure displacement of the monitoring point; The buckling failure displacement of the monitoring point is compared with the real-time displacement alarm threshold. If the real-time displacement alarm threshold of the monitoring point is greater than the buckling failure displacement of the corresponding monitoring point, the real-time displacement alarm threshold of the monitoring point is adjusted to be less than the buckling failure displacement of the corresponding monitoring point.

7. The method according to claim 1, characterized in that The modified parameters of the finite element model include: elastic modulus of the steel roof support structure, node connection stiffness and support position of the Y-shaped column.

8. The method according to claim 1, characterized in that The method for safety monitoring of the steel roof support structure based on visual recognition further comprises the following steps: Establish a digital twin model of the steel roof support structure to obtain real-time displacement of monitoring points at the construction site; The digital twin model is used to simulate the construction plan, and the monitoring points and construction stages with construction hazards are predicted by comparing the real-time displacement alarm thresholds of the monitoring points and the real-time displacements of the monitoring points at the construction site.

9. The method according to claim 8, characterized in that The method for safety monitoring of the steel roof support structure based on visual recognition further comprises the following steps: According to the monitoring points and construction stages where construction hazards exist, optimize the construction plan or the real-time displacement alarm threshold of the monitoring points.

10. A steel roof support structure safety monitoring system based on visual recognition, characterized in that: include: Monitoring module, used to build finite element model of steel roof support structure; Selecting monitoring points on the steel roof support structure according to the nodes of the finite element model; The real-time displacement of the monitoring points is obtained through visual recognition, including: setting markers that are dynamically adaptive to the environment at the monitoring points; acquiring and identifying multiple marker images through a calibrated camera to obtain the real-time displacement of multiple monitoring points; adjusting the real-time displacement alarm threshold of the monitoring points according to the real-time working conditions; The alarm module is used to send out an alarm signal when the real-time displacement of any monitoring point exceeds the corresponding alarm threshold.

Images