High-rise flower basket plate buckle scaffold and monitoring system

Abstract

The invention relates to the technical field of scaffolds, in particular to a high-rise flower basket disc buckle scaffold and a monitoring system.The high-rise flower basket disc buckle scaffold comprises a flower basket pull rod assembly, a cantilever I-shaped beam and a disc buckle scaffold body, and the cantilever I-shaped beam is connected with a lower concrete wall; the basket pull rod assembly comprises an upper pull rod, a basket and a lower pull rod, the upper pull rod and the lower pull rod are in threaded fit with the basket, an upper connecting plate connected with an upper concrete wall is arranged at the end, away from the basket, of the upper pull rod, and a lower connecting plate connected with the cantilever I-shaped beam is arranged at the end, away from the basket, of the lower pull rod. The upper connecting plate is connected with an upper mounting plate, the upper mounting plate is connected with a pressure sensor abutting against the upper concrete wall, and a tension sensor is connected between the upper pull rod and the lower pull rod; the cantilever I-shaped beam is provided with a marker of which the code can be read; and the system further comprises a monitoring host. According to the invention, the stress condition of main components of the scaffold can be monitored in real time, and the safety performance of the scaffold is well improved.

Description

Technical Field

[0001] This invention relates to the field of scaffolding technology, specifically to a high-rise flower basket buckle scaffolding and monitoring system. Background Technology

[0002] In the construction of high-rise buildings, steel cantilever scaffolding is required, among which basket-type cantilever disc-lock scaffolding is a commonly used construction auxiliary method. Given the large scale of cantilever scaffolding erection in high-rise buildings, it has a significant impact on the safety level, construction progress, and cost control of the project. Therefore, the development of prefabricated cantilever disc-lock scaffolding with better performance for high-rise buildings has become a key issue, and its structural form and construction technology have long been a focus of attention for technical management personnel.

[0003] Currently, the utility model patent with announcement number CN217439500U discloses a prefabricated cantilevered disc-lock scaffolding structure for high-rise buildings, including a concrete structure, a cantilevered I-beam main beam, a basket brace structure, a disc-lock scaffolding frame, and a support elevation adjustment device.

[0004] Existing scaffolding achieves stable connection and protection through the structure of the components themselves. Due to the large scale of scaffolding and the huge number of components, inspection is time-consuming and labor-intensive. It is difficult to detect abnormalities such as sporadic component damage, loosening, and cracking. Although current scaffolding has a large redundancy in load-bearing and structural stress at the beginning of the design, safety hazards still cannot be ignored.

[0005] In recent years, with the development of IoT and sensor technologies, the scaffolding industry has also introduced related technologies that can monitor the status of scaffolding in real time, so as to detect safety hazards in a timely manner and improve the safety of scaffolding. Summary of the Invention

[0006] Based on the above description, the present invention provides a high-rise flower basket buckle scaffolding that can monitor the stress on the main components of the scaffolding in real time, thereby improving the safety performance of the scaffolding.

[0007] On the one hand, the technical solution of the present invention to solve the above-mentioned technical problems is as follows: a high-rise flower basket scaffolding, including a flower basket tie rod assembly, a cantilevered I-beam and a scaffolding frame, wherein the cantilevered I-beam is connected to the lower concrete wall; The flower basket pull rod assembly includes an upper pull rod, a flower basket, and a lower pull rod. The upper pull rod and the lower pull rod are both threaded into the flower basket. The end of the upper pull rod away from the flower basket is provided with an upper connecting plate that connects to the upper concrete wall, and the end of the lower pull rod away from the flower basket is provided with a lower connecting plate that connects to the cantilevered I-beam. The upper connecting plate is connected to the upper mounting plate, and the upper mounting plate is connected to a pressure sensor that abuts against the upper concrete wall. A tension sensor is connected between the upper pull rod and the lower pull rod, and the two ends of the tension sensor are connected to the upper pull rod and the lower pull rod respectively through adjustable length connecting ropes. Each cantilevered I-beam is equipped with a readable coded marker, and each cantilevered I-beam corresponds to a unique code. It also includes a monitoring host for receiving real-time data from pressure and tension sensors and for storing the coded information of the markers.

[0008] It should be understood that by arranging pressure sensors at the upper connecting plate of the basket tie rod assembly and tension sensors between the upper and lower tie rods, and with the monitoring host receiving and recording data in real time, continuous quantitative monitoring of the key force link of "upper wall bearing pressure - tie rod system bearing tension" can be achieved. Compared to the reactive approach of relying on manual inspections, this method can identify and trigger early warnings at an early stage, such as component loosening, connection failure, abnormal increase or sudden change in force. The tension sensors use ropes at both ends with adjustable lengths, ensuring that installation does not change the main load-bearing structure, adapts to different tie rod spacings, and facilitates maintenance and replacement. The cantilevered I-beams are assigned unique codes and stored by the host, achieving a one-to-one correspondence between force data and specific I-beams. This facilitates rapid location of potential hazards and the formation of traceable historical force files, thereby reducing the cost of large-scale scaffolding inspections and improving the efficiency of risk management.

[0009] Based on the above technical solution, the present invention can be further improved as follows.

[0010] Furthermore, the upper mounting plate is provided with a vertically arranged elongated connecting hole for the screw to pass through, and the detection part of the pressure sensor is pressed against the surface of the concrete wall.

[0011] Furthermore, both ends of the upper and lower pull rods are connected to connecting rings, quick-release buckles are connected to the connecting ropes at both ends of the tension sensor, and an adjusting buckle for adjusting the tension of the connecting rope is connected to at least one end of the connecting rope.

[0012] Furthermore, the identifier is a QR code, barcode, or RFID tag.

[0013] Secondly, the technical solution of the present invention to solve the above-mentioned technical problems is as follows: a monitoring system for high-rise flower basket interlocking scaffolding, applied to the above-mentioned high-rise flower basket interlocking scaffolding, comprising: The sensor node group is used to collect the force data at the corresponding position of the monitored cantilevered I-beam. Each sensor node in the sensor node group has at least a unique node code. A set of monitoring target points is set below the cantilevered I-beam. The monitoring target points are coded targets that can be identified by machine vision, and each monitoring target point has a unique target point code. The image acquisition subsystem includes multiple fixed cameras and / or cameras mounted on drones, used to acquire image data containing monitoring target points; An edge gateway is used to receive and forward force data from a group of sensor nodes. The monitoring host is used to receive the force data and image data, establish and maintain the position data of "node code - target code - spatial position", analyze the image data to obtain target displacement / attitude information, and fuse the force data and target displacement / attitude information to output monitoring results and alarm information. The user terminal is connected to the monitoring host and is used to display the monitoring results and alarm information.

[0014] It should be understood that by acquiring multi-source data—"force data from sensor nodes + machine vision displacement / attitude data from monitoring targets"—and establishing and maintaining a binding relationship between "node code—target code—spatial location" by the monitoring host, each force anomaly can be precisely correlated with the corresponding spatial component's position and displacement response. Compared to single-sensor or single-vision monitoring, the fused monitoring results can simultaneously reflect load changes and structural responses, and can more sensitively identify hidden risks such as "force anomalies but small displacements" or "displacement anomalies but insignificant force" caused by loose fasteners, slippage of connections, component deformation, or local damage, significantly reducing missed and false alarms. The combination of fixed cameras and drone inspections covers a wide range and obstructed areas, while edge gateways handle the nearest aggregation and forwarding, improving communication reliability and scalability. User terminals present results and alarms in real time, enabling rapid positioning, hierarchical handling, and closed-loop management of large-scale scaffolding.

[0015] Furthermore, the image acquisition subsystem includes: a fixed camera array, whose installation position and field of view are configured to cover multiple monitoring target points and form overlapping fields of view to support multi-view joint calculation; and / or The UAV inspection unit performs photography according to the inspection tasks issued by the monitoring host. The inspection tasks include at least the sequence of flight path points, target coverage constraints, and photography triggering conditions. The monitoring host generates the route point sequence based on location data and / or BIM data, enabling the UAV to collect image data containing at least one monitoring target point at each route point and upload it with a timestamp.

[0016] Furthermore, the monitoring host includes an image analysis module for performing target detection, target encoding / decoding, and displacement calculation on image data, and calculating the first target value based on at least the following relationship. Each monitoring target at time Displacement amount: ; in, For displacement, and The monitoring target points are respectively at time. Compared to the pixel coordinates in the image coordinate system at the initial moment. The pixel-to-actual length conversion factor is determined by the camera calibration parameters and the geometric scale of the target point; the monitoring host generates the deformation characterization quantity of the corresponding position of the cantilever I-beam based on the displacement of multiple monitoring target points.

[0017] Furthermore, the monitoring host includes a multi-source binding module for automatically binding force data and image data in the spatial location dimension. The multi-source binding module establishes the binding relationship using at least one or a combination of the following information: The mounting pose of the fixed camera and the encoding of the identifiable target points within its field of view; The timestamp and shooting pose information of the drone during shooting, and the corresponding target point code; Preset constraints for node encoding and target encoding within the same location unit.

[0018] Furthermore, the monitoring host and the edge gateway form an adaptive sampling control loop: when the monitoring host determines that the risk level has increased or detects that the displacement change rate exceeds the threshold, it sends an instruction to the fixed camera to increase the frame rate and / or sends an encrypted route re-inspection instruction to the UAV inspection unit, and sends an instruction to the sensor node group to increase the sampling frequency and reporting frequency; when communication is interrupted, the edge gateway and / or sensor nodes locally cache the data and retransmit the data packets with timestamps after communication is restored.

[0019] Furthermore, the monitoring host includes a risk assessment and graded alarm module, used to calculate risk indicators based on force data and displacement data and trigger alarms, wherein the risk indicators at least meet the following requirements: ; in, This represents the real-time force value. As the reference force value, To allow for a force threshold, This refers to the maximum displacement of the monitored target point within the same spatial location unit. To allow displacement threshold, For weighting coefficients; when When preset conditions are met, the monitoring host pushes alarm information containing target code, spatial location, risk level, trigger data and timestamp to the user terminal, and automatically generates a review task to drive the drone to re-inspect or the designated fixed camera to re-shoot.

[0020] Compared with the prior art, the technical solution of this application has the following beneficial technical effects: 1. Online quantification and traceable positioning of key stress links: Pressure and tension sensors are installed at the basket tie rod assembly to collect and record the key stress path in real time and for a long time: "pressure against the wall - tension on the tie rod - stress on the cantilever beam"; at the same time, each cantilever I-beam is equipped with a unique readable code (QR code / barcode / RFID) and stored by the monitoring host to achieve a one-to-one correspondence between "data and component". This can shift the detection of potential problems such as loosening, cracking, and connection failure from "relying on manual inspections for post-event discovery" to "online early warning + precise positioning", significantly reducing the cost of large-scale scaffolding inspections and improving the efficiency of handling them; 2. Multi-source data spatial binding and multi-view joint calculation of deformation / displacement: Through the fusion monitoring of "sensor node group force data + monitoring target point machine vision displacement / attitude data", the monitoring host establishes and maintains position data of "node code - target code - spatial position", realizing automatic binding and joint analysis of force and displacement within the same position unit. The overlapping field of view of the fixed camera supports multi-view calculation, and the UAV fills in the occlusion and blind spots according to the flight path point sequence and coverage constraints; the displacement is calculated according to the pixel coordinate difference and combined with the calibration conversion coefficient, and then the deformation characterization of the cantilever beam is generated. Compared with single sensor or single vision, the false alarm / false alarm rate is lower and the positioning is more accurate. 3. Risk-driven adaptive sampling, disconnection buffering and retransmission, and hierarchical closed-loop alarms; the monitoring host and edge gateway form an adaptive sampling control loop: when the risk increases or the displacement change rate exceeds the threshold, the camera frame rate is automatically increased, the drone re-inspection route is encrypted, and the sampling / reporting frequency of sensor nodes is increased; when the risk decreases, sampling is reduced to save energy and reduce load. When communication is interrupted, the gateway / node locally caches and retransmits timestamped data after recovery to ensure continuous reconstruction and traceability. The risk assessment module calculates indicators based on the ratio of force deviation to displacement and pushes alarms hierarchically, automatically generating review tasks to drive re-shooting / re-inspection, realizing closed-loop management of "monitoring-alarm-review". Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall structure of a high-rise flower basket buckle scaffolding provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the connection structure between the flower basket tie rod assembly and the concrete wall in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the structure of the flower basket pull rod assembly according to Embodiment 1 of the present invention; Figure 4 This is a block diagram of the overall structure of the monitoring system according to Embodiment 2 of the present invention; Figure 5 This is a schematic diagram of the on-site layout of Embodiment 2 of the present invention; Figure 6This is a schematic diagram illustrating the binding relationship between the coding system of Embodiment 2 of the present invention and the "node coding - target coding - spatial location". Figure 7 This is a flowchart of image acquisition and task organization in Embodiment 2 of the present invention; Figure 8 This is a flowchart of the image analysis and displacement calculation process in Embodiment 2 of the present invention; Figure 9 This is a schematic diagram of the force-displacement fusion, risk assessment and classification alarm, and closed-loop control in Embodiment 2 of the present invention.

[0022] Reference numerals: 1. Flower basket pull rod assembly; 11. Upper pull rod; 12. Flower basket; 13. Lower pull rod; 14. Upper connecting plate; 15. Lower connecting plate; 16. Pressure sensor; 17. Tension sensor; 18. Connecting rope; 2. Cantilevered I-beam; 21. Marker; 3. Upper mounting plate; 31. Long strip connecting hole; 4. Upper concrete wall; 5. Lower concrete wall. Detailed Implementation

[0023] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings, which illustrate embodiments of the present application. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of this application will be thorough and complete.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0025] Example 1: refer to Figures 1-3 A high-rise flower basket scaffolding includes a flower basket tie rod assembly, a cantilevered I-beam, and a scaffolding frame. The cantilevered I-beam is connected to a lower concrete wall. During the pouring of the lower concrete wall, bolt sleeves are pre-embedded in it. High-strength connecting bolts are used to penetrate the bolt sleeves to fix the cantilevered I-beam to the lower concrete wall. A positioning buckle is fitted on the cantilevered I-beam, and a positioning post extends upward from the positioning buckle. The upright of the scaffolding frame is fitted onto the positioning post.

[0026] The basket pull rod assembly includes an upper pull rod, a basket, and a lower pull rod. Both the upper and lower pull rods are threaded into the basket. The end of the upper pull rod away from the basket is equipped with an upper connecting plate that connects to the upper concrete wall, and the end of the lower pull rod away from the basket is equipped with a lower connecting plate that connects to the cantilevered I-beam. Bolt sleeves are also pre-embedded in the upper concrete wall. The upper connecting plate is connected to the upper concrete wall by using a high-strength screw that passes through the bolt sleeves pre-embedded in the upper concrete wall. The upper connecting plate is connected to the upper mounting plate. The upper mounting plate is provided with a vertically arranged elongated connecting hole for the screw to pass through. The screw passes through the upper mounting plate, the upper connecting plate and the upper concrete wall. The upper mounting plate is inverted L-shaped. A pressure sensor that abuts against the upper concrete wall is connected to the horizontal section of the upper mounting plate. The detection part of the pressure sensor is pressed against the surface of the concrete wall. Furthermore, the pressure sensor in this embodiment includes a disc-shaped pressure detection module, a power supply module, and an Internet of Things (IoT) module. The power supply module provides power to the pressure detection module and the IoT module, and the IoT module is used to transmit the mechanical data detected by the pressure detection module.

[0027] A tension sensor is connected between the upper and lower pull rods. The two ends of the tension sensor are connected to the upper and lower pull rods respectively via adjustable-length connecting ropes. Each end of the upper and lower pull rods has a connecting ring. Quick-release buckles are attached to the connecting ropes at both ends of the tension sensor. At least one end of the connecting rope has an adjusting buckle for adjusting the tension of the connecting rope. The adjusting buckle is an "8-shaped adjusting buckle". The quick-release buckle allows for easy and quick installation of the tension sensor between the upper and lower pull rods, and the overall connection length of the connecting rope can be adjusted by the adjusting buckle. Furthermore, the tension sensor in this embodiment includes a tension detection module, a power supply module, and an Internet of Things (IoT) module. The power supply module provides power to the tension detection module and the IoT module, and the IoT module is used to transmit the mechanical data detected by the tension detection module.

[0028] Each cantilevered I-beam is equipped with a readable coded identifier. Each cantilevered I-beam corresponds to a unique code, which can be a QR code, barcode, or RFID tag.

[0029] It also includes a monitoring host for receiving real-time data from pressure and tension sensors and for storing the coded information of the markers.

[0030] Example 2: Reference Figures 4-9 A monitoring system for high-rise basket-lock scaffolding is disclosed, applied to the high-rise basket-lock scaffolding described in Example 1. This system is used to collaboratively monitor the stress state and displacement / attitude changes at corresponding locations of cantilevered I-beams, and to achieve risk assessment, tiered alarms, and closed-loop verification. The monitoring system includes: a sensor node group, a monitoring target point group, an image acquisition subsystem, an edge gateway, a monitoring host, and a user terminal.

[0031] 1. System Structure and Coding System (1) Sensor node group: The sensor node group includes at least one sensor node. Each sensor node has at least a unique node code, which is used to collect stress data related to the corresponding position of the monitored I-beam and generate a data packet with a timestamp for uploading. The stress data may include the output value of the pressure sensor, the output value of the tension sensor, or a combination thereof. The data packet may include the node code, sampling timestamp, sensor value, sequence number, and verification field to support missing packet detection and out-of-order reordering.

[0032] (2) Monitoring target group: The monitoring target group is set under the cantilevered I-beam. The monitoring target is a coded target that can be recognized by machine vision, and each monitoring target has a unique target code. The target can be a planar high-contrast coded target, and a feature structure (such as corner point / dot array) is set for stable extraction to simultaneously meet the needs of "coding recognition" and "displacement calculation".

[0033] (3) Image acquisition subsystem. The image acquisition subsystem includes multiple fixed cameras and / or cameras mounted on UAVs, used to acquire and upload image data containing the monitored target points. The image data must carry at least a timestamp; when using a UAV, it may also carry shooting pose information (e.g., heading, pitch, gimbal angle, position coordinates, etc.).

[0034] (4) Edge gateway: The edge gateway is used to receive the force data uploaded by the sensor node group and forward it to the monitoring host; optionally, the edge gateway performs local caching, protocol conversion, encryption and preliminary verification of the data.

[0035] (5) Monitoring host: The monitoring host is used to receive force data and image data, establish and maintain the position data of “node code - target code - spatial position”, analyze the image data to obtain target displacement / attitude information, and fuse the force data and target displacement / attitude information to output monitoring results and alarm information, and at the same time form a verification closed loop task.

[0036] (6) User terminal: The user terminal is connected to the monitoring host for displaying monitoring results and alarm information, and is used for initial table creation, verification of task execution and result backfilling.

[0037] 2. Image acquisition methods and task organization (1) Fixed camera array method: The fixed cameras are installed on the edge of the floor, structural beams or temporary supports. Their installation position and field of view are configured to cover multiple monitoring target points and form at least partially overlapping fields of view to support multi-view joint calculation and occlusion redundancy. The monitoring host can save the installation pose and calibration parameters of the fixed cameras and maintain the set of visible cameras for the target points.

[0038] (2) UAV inspection method: The UAV inspection unit performs shooting according to the inspection task issued by the monitoring host. The inspection task includes at least: flight path sequence, target coverage constraint and shooting trigger condition. Among them, the monitoring host can generate flight path sequence based on location data and / or BIM data, so that the UAV can collect image data containing at least one monitoring target at each flight path and upload it with timestamp; the shooting trigger condition may include point trigger, gimbal attitude constraint trigger or target entry into field of view ratio reaching threshold trigger.

[0039] 3. Automatic binding of multi-source data and maintenance of location data: To achieve the fusion of force data and displacement data at the same spatial location, the monitoring host is equipped with a multi-source binding module to establish and update the location data of "node code - target code - spatial location". The binding relationship is established using at least one of the following information or a combination thereof: a) the installation pose of the fixed camera and the identifiable target code within its field of view; b) the timestamp and shooting pose information of the UAV during shooting and its corresponding target code; c) the preset constraints of node code and target code within the same location unit.

[0040] To ensure feasibility, this embodiment provides a set of optional implementation procedures: Initial Table Building: When installing sensor nodes and monitoring targets, construction personnel scan the I-beam markers and target codes using user terminals, and select the corresponding node code (or automatically read the node code from near-field communication), forming an initial mapping of "I-beam code - node code - target code"; this is then written into the spatial location field in conjunction with BIM or measurement data. Dynamic Maintenance: When the UAV inspection obtains images from a new perspective, the monitoring host updates the target visibility and scale conversion parameters based on "shooting pose + target code" and writes back the location data. Consistency Verification: During system operation, when inconsistencies such as "significant force changes while displacement remains near zero for a long time" or "significant displacement while force remains near zero for a long time" occur, the monitoring host triggers a binding verification, outputting a verification prompt or adjusting binding candidates based on a set of neighboring target points.

[0041] 4. Image Analysis and Displacement Calculation: The monitoring host includes an image analysis module for performing target detection, target encoding / decoding, and displacement calculation on image data. This can be achieved through the following steps: S1) Distortion correction and ROI cropping of the input image; S2) Detecting target candidate regions and decoding to obtain target codes, while simultaneously extracting the pixel coordinates of target feature points; S3) Calculating the pixel displacement of the target at time t relative to the initial time and converting it into the actual displacement.

[0042] The displacement satisfies: ;in, For displacement, and These are the pixel coordinates of the target point at time t and at the initial time, respectively. This is the pixel-to-actual-length conversion factor.

[0043] It can be determined by the known geometric scale of the target point: for example, the actual distance between the target feature points is... The pixel spacing measured in the image is Then it is acceptable. When using a drone and the shooting distance changes, It can also be dynamically updated based on shooting position or distance estimation to reduce scale drift.

[0044] The monitoring host can aggregate the displacement of multiple target points under the same I-beam to generate deformation characteristics of the corresponding position of the I-beam (e.g., taking the maximum value, average value, or weighted by position) for risk assessment.

[0045] 5. Force-displacement fusion, risk assessment, and graded alarm: The monitoring host fuses force data and displacement / attitude information within the same spatial unit, outputs monitoring results, calculates risk indicators based on the fusion results, and triggers alarms. Risk indicators must meet the following conditions: ; in, This represents the real-time force value. The reference stress value (e.g., the stable average value after installation or the value confirmed manually). To allow for a force threshold, This represents the maximum displacement of the target point within the same spatial location unit. To allow displacement threshold, These are the weighting coefficients.

[0046] when When preset conditions are met, the monitoring host pushes alarm information to the user terminal. The alarm information includes at least: target code, spatial location, risk level, trigger data and timestamp; and automatically generates a review task to drive the drone to re-inspect or the designated fixed camera to re-shoot.

[0047] 6. Adaptive sampling control loop and disconnection buffer retransmission: The monitoring host and edge gateway constitute an adaptive sampling control loop. When the monitoring host determines that the risk level has increased or detects that the displacement change rate exceeds the threshold, it issues an instruction to increase the sampling strategy, including: issuing an instruction to increase the frame rate to the fixed camera; and / or Issue encrypted route re-inspection instructions to the UAV inspection unit; and / or Send instructions to the sensor node group to increase the sampling frequency and reporting frequency.

[0048] When the risk level decreases and the data stabilizes, a downsampling command is issued to reduce communication and energy consumption. When communication is interrupted, the edge gateway and / or sensor node caches the data locally and retransmits the data packets with timestamps after communication is restored to ensure that the monitoring host can perform continuous reconstruction and traceability analysis.

[0049] As a preferred embodiment, 1. System configuration and coding system, (1) Sensor node group, in this embodiment, the sensor node group includes at least one sensor node, which is deployed on the force transmission path corresponding to the cantilever I-beam, and is used to collect the force data at that location and report it to the monitoring host.

[0050] 1) Node Composition: Each sensing node may include: a sensing acquisition unit, an encoding and storage unit, a clock / synchronization unit, a communication unit, and a power supply unit. The sensing acquisition unit is used to connect to pressure sensors and / or tension sensors and perform analog-to-digital conversion; the encoding and storage unit is used to store the unique node code of the node (which can be burned into a chip ID, EEPROM, or security module); the clock / synchronization unit is used to generate sampling timestamps; the communication unit is used to establish a data connection with the edge gateway (wireless or wired); the power supply unit can be battery-powered, externally powered, or powered by energy harvesting.

[0051] 2) Node coding and mapping relationship: Each sensing node has at least one unique node code (NodeID) to achieve traceability of "data-device-spatial location". The NodeID can be in string or integer coding form, and the coding rule can be a combination of project number + floor number + I-beam number + node sequence number, or a randomly generated and non-repeating UUID code; the coding rule can be configured on the monitoring host side and does not constitute a limitation on the protection scope.

[0052] 3) Force data content, wherein the force data includes at least the pressure value P and / or the tension value T. Preferably, in addition to uploading real-time values, the sensing node can also upload sensor status information, such as: range identifier, temperature compensation amount, zero drift amount, offline marker, etc., to improve engineering usability.

[0053] 4) Data packet format: The reporting data packet generated by the sensor node shall include at least: Node ID; sampling timestamp ts (local clock or gateway synchronization clock); sensor value (P and / or T); and optionally include: sequence number seq (for packet loss detection and out-of-order reordering); checksum field rc (for transmission integrity verification); and power / signal strength status (for maintenance). The timestamp ts is used for alignment with the timeline of the image data; seq and rc are used for data quality control in scenarios with unstable communication.

[0054] 5) Sampling and reporting strategy: Sensor nodes can perform sampling according to the sampling parameters issued by the monitoring host, including sampling frequency. Reporting frequency And window aggregation methods (e.g., moving average, peak hold). In the default mode, nodes are displayed in a fixed manner. Sampling and fixing Report; in the risk-enhanced mode, the node increases its risk level according to instructions. and .

[0055] (2) Monitoring target group: In this embodiment, the monitoring target group is set below the cantilevered I-beam to provide visually identifiable codes and stable geometric features to support displacement / attitude measurement.

[0056] 1) Target structure and coding region design: Each monitoring target is a planar coding target, which includes at least a coding region and a feature region.

[0057] The encoding area is used to carry the target ID (e.g., QR code / custom black and white code / AprilTag type encoding). Feature regions are used to provide stably extracted feature points (such as corner points, dot arrays, or high-contrast contours) to support subpixel-level localization.

[0058] 2) Target ID Coding and Uniqueness: Each monitoring target has a unique target ID (TargetID). The TargetID can be mapped to the I-beam identification code and node code to construct location data in the format of "node code—target ID—spatial location". Preferably, the TargetID can be consistent with the construction sequence, for example, generated by floor—beam segment—installation sequence number, facilitating on-site management and verification.

[0059] 3) Target mounting and attitude constraints: The target can be fixed to the lower flange of the I-beam or its associated mounting plate using bolts, magnetic bases, clamps, or adhesive structures. To reduce ambiguity in attitude calculation, preferably, the target plane is parallel to the bottom surface of the I-beam or has a recordable initial attitude; and the initial reference image / initial pixel coordinates are recorded during the initial table creation. .

[0060] 4) Target group layout principle: Within the monitoring range corresponding to the same I-beam, multiple monitoring target points can be set to cover key stress and deflection locations (such as the outer end, mid-span, and wall end), so that the monitoring host can aggregate the displacement of multiple target points to form a deformation characterization quantity.

[0061] (3) Image acquisition subsystem: The image acquisition subsystem is used to acquire image data containing the monitoring target and upload it to the monitoring host, providing input for displacement / attitude calculation.

[0062] 1) Composition: The image acquisition subsystem includes multiple fixed cameras and / or cameras mounted on drones. Fixed cameras are used for continuous or high-frequency acquisition, while drones are used for inspection, filling in gaps, and verification; both can be used individually or in combination.

[0063] 2) Minimum necessary fields for image data: Each frame of image data should carry at least the following: acquisition device identifier CamID / DroneID, timestamp ts_img, and image body; when using a drone, it is preferable to carry the shooting pose (e.g., position coordinates, heading / pitch / roll, gimbal angle) for scale conversion parameter updates and binding maintenance.

[0064] 3) Camera calibration and parameter management: Preferably, the monitoring host stores the intrinsic parameters and distortion parameters of the fixed camera, and can also store the installation pose; the intrinsic parameters of the UAV camera can be pre-calibrated or stored permanently according to the equipment model. These parameters are used for distortion correction, target localization, and pixel-to-actual length conversion factors. The determination or update.

[0065] 4) Triggering and reporting strategies: Fixed cameras can collect and report data at a set frame rate; UAVs can trigger data collection based on flight path points, attitude constraints, or target entry into the field of view. The monitoring host can issue adjustment commands (increase frame rate / enhance re-inspection) according to the risk level, forming an adaptive closed loop.

[0066] (4) Edge gateway: The edge gateway is used to aggregate sensor data on the field side and forward it to the monitoring host, while providing data preservation capabilities in the event of network outage.

[0067] 1) Interface and connection method: The edge gateway should include at least an access interface (wireless or wired) for communicating with sensor nodes and an uplink interface (e.g., Ethernet / cellular network) for communicating with the monitoring host. The gateway can be deployed near the temporary power distribution box on the floor or in a safe area to improve signal coverage and ease of maintenance.

[0068] 2) Local caching and retransmission mechanism: When the monitoring host is unreachable or the link is interrupted, the edge gateway will cache the received data packets in the order of timestamps; after communication is restored, the packets will be retransmitted in the order of timestamps and carry the original seq and ts fields, so that the monitoring host can complete continuous reconstruction and traceability analysis.

[0069] 3) Protocol conversion and security: The edge gateway can convert the short-range protocols of the sensor nodes and encrypt and verify the integrity of the uplink data; it can also perform basic anomaly removal (such as over-range marking) without affecting the retention of the original data.

[0070] (5) Monitoring host: The monitoring host is the core of the system and is used for multi-source data reception, location data modeling and maintenance, image analysis, fusion calculation, risk assessment alarm and review closed loop.

[0071] 1) Module division: The monitoring host may include: data access module, location data maintenance module (multi-source binding module), image analysis module, fusion analysis module, adaptive sampling control module, risk assessment and graded alarm module, task management module, and data storage module.

[0072] 2) Location data structure: The monitoring host establishes and maintains a location data table, which includes at least the following fields: Node ID; Target ID; Spatial location (Loc, which can be BIM component ID, floor / axis / elevation combination, or 3D coordinates); and optionally includes: Beam ID (from the I-beam marker); VisCamSet (set of visible cameras); Pixel-to-actual length conversion parameter. It also includes the source markers (fixed camera calibration / UAV pose update / manual calibration). This table enables the fusion analysis of force data and target displacement / attitude information within the same spatial unit.

[0073] 3) Time alignment and fusion entry: The monitoring host aligns the force data and image data based on the timestamp. It can use the "nearest neighbor matching" or "time window aggregation" method to align and enter the fusion analysis module; and write the fusion results into the monitoring results table for display and alarm use.

[0074] 4) Alarm and review task driven: When the monitoring host outputs an alarm, it simultaneously generates a review task and sends it to the control terminal of the drone or fixed camera. The review task is bound to the spatial location (Loc) and target code (TargetID) to facilitate closed-loop verification and record keeping.

[0075] (6) User terminal: The user terminal is used for on-site table creation, operation and maintenance management and alarm response. It is the execution carrier of "digital handover and review".

[0076] 1) Communication and display: The user terminal communicates with the monitoring host to display monitoring results, risk levels and alarm information in real time, and can retrieve historical curves and event records of I-beams / target points / nodes by spatial location.

[0077] 2) Initial construction table creation: The user terminal is used for initial construction table creation: scanning the I-beam identification code, scanning the target point code, and selecting or reading the node code to form an initial mapping relationship; at the same time, spatial location information such as floor, axis, and component number can be entered / selected and written into the location data table.

[0078] 3) Review task execution and backfilling: When the monitoring host issues a review task, the user terminal prompts the review location and triggering reason; after the review is completed, the user terminal can backfill the review conclusion (e.g., reinforced / reset / false alarm reason) and upload on-site photos or re-photograph results for updating the baseline value or threshold strategy.

[0079] Preferably, in the image acquisition method and task organization, in this embodiment, the image acquisition subsystem is used to acquire image data containing monitoring target points, which serves as input for the monitoring host to perform target point identification, displacement / attitude calculation, and fusion analysis with force data. The image acquisition subsystem may include multiple fixed cameras and / or cameras mounted on a drone, which can be used individually or in combination: the fixed camera array is used for continuous or quasi-continuous acquisition of monitoring target points, and the drone patrol is used to supplement acquisition of blind spots and obstructed areas of the fixed cameras, and can also serve as a verification method after a risk alarm.

[0080] Specifically, the fixed camera array is installed on the edge of the floor, structural beams, the edge of the core tube, or temporary supports. The monitoring host plans and configures the installation position and field of view of the fixed cameras, so that the field of view of a single fixed camera covers multiple monitoring target points to improve coverage efficiency, and multiple fixed cameras form at least partially overlapping fields of view for the same monitoring target point to provide occlusion redundancy and support multi-view joint calculation. To facilitate construction error compensation and subsequent maintenance, the fixed cameras can be finely adjusted in pitch and azimuth angles through adjustable pan-tilt units or adjustable supports, thereby ensuring that the target point is located in the effective image area. In this embodiment, each fixed camera has a unique device identifier CamID. The monitoring host maintains a corresponding parameter set for CamID. The parameter set includes at least camera intrinsic parameters and distortion parameters, and optionally includes installation pose information and a field of view coverage list. The monitoring host uses the above parameters to perform distortion correction, target point localization, and pixel-to-actual length conversion parameter updates on the acquired images. At the same time, the installation pose of the fixed camera and the identifiable target point codes within its field of view are used as one of the input information for subsequent multi-source binding. To adapt to various working conditions such as the addition or reduction of target points, changes in obstruction, or adjustments to the support structure, the monitoring host also maintains a set of visible cameras (VisCamSet) for each monitoring target point. The VisCamSet can be generated through the configuration confirmation method during the initial table creation during construction, or it can be updated in a learning manner based on the successful decoding statistics of a certain CamID-TargetID during system operation. It can also perform geometric visibility verification by combining the camera pose and the spatial position of the target point, thereby realizing automatic maintenance of coverage relationships.

[0081] Fixed cameras perform data acquisition and reporting according to the acquisition strategy issued by the monitoring host. This acquisition strategy includes at least frame rate, compression quality or bit rate, upload method, and acquisition period. In normal monitoring mode, fixed cameras acquire data at a basic frame rate. In high-risk mode, the monitoring host can issue instructions to a specified CamID to increase the frame rate and / or increase the resolution to enhance the ability to capture displacement changes. When communication bandwidth is limited or the network is congested, fixed cameras can also choose to upload cropped images containing the target area or upload target decoding results and feature point coordinates to reduce bandwidth usage without affecting the displacement calculation of the monitoring host. The image data uploaded by fixed cameras includes at least the CamID, image timestamp, and image body. The monitoring host indexes camera parameters based on the CamID and enters the image analysis process. When a target fails to decode continuously or feature point extraction fails under a certain CamID, the monitoring host can call images from other cameras in its VisCamSet as compensation input. If these are still unavailable, the target is marked as visually missing, and a drone re-shoot or on-site verification is triggered to ensure monitoring continuity.

[0082] Furthermore, the UAV inspection unit performs filming according to the inspection tasks issued by the monitoring host. These inspection tasks include at least a sequence of waypoints, target point coverage constraints, and filming trigger conditions, and are used to quickly acquire target points when fixed camera coverage is insufficient or rapid verification is required. In this embodiment, the UAV has a unique device identifier (DroneID), and each inspection task has a task identifier (TaskID). The monitoring host records the task issuance time, execution time period, and target point set for traceability and record keeping. The monitoring host generates a sequence of waypoints (WayPoints) based on location data and / or BIM data, which consists of "node code—target code—spatial location." Each waypoint includes at least the waypoint's location coordinates, elevation or height constraints, recommended gimbal attitude, and the corresponding target point set. The waypoint sequence can be generated according to a strategy of "prioritizing high-risk locations and then general locations" to shorten the high-risk verification delay. To ensure that the images acquired by the UAV can be decoded and used for displacement calculation, the target coverage constraints may include the minimum pixel size constraint of the target point within the field of view, the proportion of the visible area of ​​the target point, the shooting distance range constraint, and the gimbal attitude and line-of-sight angle constraint. These constraints can be preset by the monitoring host or adaptively adjusted based on historical successful identification statistics. The shooting triggering conditions include at least one or more combinations of point-to-point triggering, attitude triggering, and frame-entry triggering: that is, shooting is triggered when the UAV reaches the flight path point and the position error is less than a threshold; or shooting is triggered when the attitude of the UAV and gimbal meets a preset attitude window; or shooting is triggered when the real-time preview detects a candidate area for the target point and the proportion of the target point in the frame reaches a threshold. Multiple frames can be continuously acquired at the same flight path point and the best frame can be uploaded to resist sample failure caused by wind disturbance, positioning errors, or short-term occlusion.

[0083] The data uploaded by the drone includes at least DroneID, TaskID, image timestamp, and image data. Preferably, it also carries shooting pose information and waypoint indexes so that the monitoring host can associate the image with the target point set and update the scale conversion parameters and binding relationships. Inspection tasks can be divided into routine inspection tasks and review inspection tasks. The review inspection task is triggered by an alarm and prioritizes covering the target points within the trigger spatial location unit. The monitoring host can adjust the task priority, waypoint density, and number of frames collected per point according to the risk level to achieve a task organization mode of "the higher the risk, the denser the collection". When the drone fails to meet the trigger conditions or fails to decode at a waypoint, it can perform remedial strategies such as attitude fine-tuning and reshooting, inserting nearby compensation waypoints to change the perspective, or marking it as a drone missing measurement and transferring it to a fixed camera for reshooting / manual review. The above execution process is recorded in the task log of TaskID for subsequent review and parameter optimization. By combining the aforementioned fixed camera array method with the UAV inspection method, this embodiment can stably acquire monitoring target image data under complex obstruction and communication fluctuation conditions at high-rise construction sites, providing a reliable data foundation for subsequent multi-source binding, displacement calculation, force-displacement fusion analysis, and risk assessment and alarm.

[0084] Preferably, 3. Automatic binding of multi-source data and maintenance of location data; In this embodiment, to achieve alignment and fusion of force data and displacement / attitude data in the same spatial dimension, the monitoring host is equipped with a multi-source binding module to establish and maintain location data (also known as a location mapping table or location data table) of "node code - target code - spatial location". The location data is used to characterize: which spatial location unit corresponds to the force data collected by the sensing node, and how the monitoring target point under that spatial location unit is identified and calculated in the image, so that the monitoring host can fuse and analyze the force data and target displacement / attitude information of the same spatial location and output monitoring results and alarm information.

[0085] In this embodiment, when the multi-source binding module establishes a binding relationship, it utilizes at least one or a combination of the following information: First, the installation pose of the fixed camera and the identifiable target code within its field of view. That is, the monitoring host uses the CamID index of the fixed camera to determine its installation pose and calibration parameters, and obtains the set of successfully decoded target codes within the camera's field of view through image analysis, thereby forming a "CamID-target code" visibility relationship. Second, the timestamp and shooting pose information of the UAV during inspection and shooting, and its corresponding target code. That is, the monitoring host uses the TaskID, timestamp, and shooting pose uploaded by the UAV to associate the identified target code with the spatial perspective of the shooting, thereby inferring the visibility and scale conversion parameters of the target in the spatial dimension and using it to update the position data. Third, the preset constraints of node codes and target codes within the same position unit. That is, during the construction and table building stage, it is pre-determined that "a certain node code corresponds to a certain I-beam or a certain position unit, and which target codes are included in the position unit", so that subsequent fusion analysis can directly complete the data association within the same position unit. By combining the above information, the multi-source binding module can maintain a stable binding relationship even when there is occlusion, camera angle adjustment, or change in inspection perspective, thus avoiding mapping drift caused by relying on only a single information source.

[0086] To ensure the effective implementation of this binding relationship, the monitoring host preferably maintains a location data table, which includes at least: NodeID, TargetID, and Loc. The Loc can be expressed using BIM component ID, floor / axis / elevation combination coding, or three-dimensional coordinates. The location data table may also optionally include: BeamID (I-beam code), VisCamSet (visible camera set), and a pixel-to-actual length conversion factor related to the target point. The data includes its source markers (fixed camera calibration / UAV pose update / manual calibration), as well as fields such as mapping confidence and update time, for subsequent consistency verification and maintenance traceability. The monitoring host uses this location data table as the entry index for multi-source data fusion: when force data encoded by a node is received, the corresponding spatial location unit (Loc) is located based on the NodeID, and the set of TargetIDs associated with that location unit is retrieved; when image data containing a TargetID is received, the Loc to which it belongs is located based on the TargetID, and the NodeID under the same Loc is retrieved, thereby achieving automatic alignment of "force-displacement" in the spatial location dimension.

[0087] During the initial table setup phase of construction, after installing sensor nodes and monitoring targets, construction personnel execute the initial mapping establishment process through the user terminal: First, they scan the markers of the cantilevered I-beam to obtain the I-beam code BeamID, and simultaneously scan the monitoring target to obtain the target code TargetID; second, they select the node code NodeID corresponding to the I-beam on the user terminal, or read the node code of the sensor node through near-field communication to reduce manual input errors; subsequently, the monitoring host generates the initial mapping relationship of "BeamID—NodeID—TargetID" and writes it into the location data table. Further, the user terminal can synchronously write the component ID of the I-beam component in the BIM system or its positioning information under the floor / axis into the spatial location Loc field; if the project does not use BIM, floor / axis / elevation information can be written into the Loc field through total station measurement, laser ranging, or manual input, thus giving the mapping relationship spatial semantics, facilitating subsequent querying and alarm location by floor, component, or area. To improve the reliability of the initial table creation, the user terminal can also guide construction personnel to perform a "table creation confirmation shot" of the target point, and the monitoring host records the initial pixel coordinates of the target point accordingly. The corresponding CamID or DroneID provides a benchmark for subsequent displacement calculation and visibility maintenance.

[0088] During system operation, the multi-source binding module executes a dynamic maintenance process to address changes in viewpoint and occlusion. Specifically, when a UAV inspection acquires a new viewpoint image, the monitoring host uses the image's timestamp and shooting pose, combined with the target point encoding TargetID set obtained from image analysis, to update the visibility relationships of the target points and write it back to the location data table. For example, it updates the set of visible cameras or the set of visible viewpoints corresponding to the TargetID. Simultaneously, it updates the pixel-to-actual length conversion coefficient of the target points based on the UAV's shooting distance or pose estimation. The system includes source markers to ensure consistent displacement calculations for the same target point across different shooting angles. For fixed camera arrays, when the field of view of a CamID changes due to bracket adjustments, the monitoring host can automatically reconstruct its field of view coverage list based on the target point decoding results within a certain time window, and update the VisCamSet of the relevant TargetID accordingly, avoiding manual reconfiguration.

[0089] Regarding consistency verification, the monitoring host preferably performs self-checks on the binding relationships of the location data table periodically or triggered by events. When inconsistencies such as "significant force changes with displacement remaining close to zero for a long period" or "significant displacement with force remaining close to zero for a long period" occur, the monitoring host triggers the binding verification process: On the one hand, it performs correlation tests or trend consistency judgments on the force curve and displacement curve within the spatial location unit (Loc), and reduces the current mapping confidence accordingly; on the other hand, it constructs a candidate mapping set based on the target point set and node set of neighboring spatial location units, for example, searching for candidate TargetIDs under the same I-beam encoding BeamID or adjacent beam segments, and re-scoring based on camera visibility, UAV pose perspective, and historical decoding success rate to generate binding candidate adjustment suggestions. If the confidence is lower than the threshold, the monitoring host outputs a verification prompt to the user terminal and issues a verification task, guiding on-site personnel to check whether there are mismatches in node installation positions, target point installation positions, or encoding scans; after the verification is completed, the user terminal fills the confirmation result back into the monitoring host, and the monitoring host updates the location data table and records the change log accordingly to form a traceable binding maintenance closed loop. Through the above-mentioned initial table creation, dynamic maintenance and consistency verification mechanism, this embodiment can maintain the binding relationship of "node code - target code - spatial location" stably for a long time in complex construction sites, providing a reliable data alignment basis for subsequent force-displacement fusion analysis, risk assessment and alarm.

[0090] Preferably, in embodiment 4, image analysis and displacement calculation, the monitoring host is equipped with an image analysis module to perform target detection, target encoding and decoding, and displacement / attitude calculation on the image data uploaded by the fixed camera array and / or the UAV inspection unit. This outputs the displacement of each monitored target point over time and further generates deformation representations of the corresponding positions of the I-beam for subsequent stress-displacement fusion analysis and risk assessment alarms. To ensure the algorithm's feasibility under complex lighting, vibration, and occlusion conditions at the construction site, the image analysis module preferably divides the processing flow into image preprocessing, target localization and decoding, feature point extraction and tracking, displacement conversion, and quality control stages, outputting intermediate results and quality indicators for traceability at each stage.

[0091] Specifically, after receiving image data, the monitoring host first performs image preprocessing. Based on the CamID or DroneID indexed in the image data, the monitoring host performs distortion correction on the input image to reduce the impact of lens distortion on the accuracy of feature point localization. Subsequently, to reduce computational load and improve real-time performance, the monitoring host performs ROI cropping on the image: firstly, the ROI can be generated based on the set of visible cameras for the target point in the location data table and the predicted pixel positions of the target point in historical frames; secondly, when the system initially identifies the target point or when occlusion occurs, a full-frame search can be used to recover the target point and re-establish the ROI. Through this "prediction-cropping-recovery" mechanism, image analysis can maintain stable frame processing efficiency even with high-resolution image input.

[0092] During the target localization and encoding / decoding stage, the monitoring host detects candidate target regions within the ROI or the entire frame and performs encoding / decoding on these regions to obtain the target code TargetID. To improve decoding robustness, target detection can preferably be based on the target's high-contrast contour, regular geometric structure, or localization features of the coded region. Perspective correction is then applied to the candidate regions before decoding. After successful decoding, the monitoring host further extracts the pixel coordinates of feature points for displacement calculation within the target feature region. These feature points can be target corners, center points, or other stable geometric key points. To improve sub-pixel localization accuracy, the monitoring host can perform sub-pixel optimization on the feature points and output feature point extraction quality indicators, such as sharpness scores, corner response values, or reprojection errors, for subsequent quality control and anomaly removal.

[0093] During the displacement calculation stage, the monitoring host uses the pixel coordinates of the target point at the initial time as a reference and calculates the pixel displacement vector of the target point at time t. Preferably, the monitoring host records the initial reference frame when the system first builds the table or when the target point is first stably identified, and marks the pixel coordinates of the target point corresponding to the reference frame as... When events such as target re-attaching, camera reinstallation, or significant viewpoint switching occur, the monitoring host can trigger a baseline reset via event flags to ensure the interpretability of the displacement calculation. The displacement calculation satisfies the following relationship: ;in, Let be the displacement of the i-th monitoring target point at time t. and These are the pixel coordinates of the monitored target point in the image coordinate system at time t and the initial time, respectively. This is the pixel-to-actual-length conversion factor. Therefore, the image analysis module can convert pixel-level displacement into actual length displacement, providing a unified dimension for subsequent fusion with force data.

[0094] To ensure full disclosure and ensure It can be determined that this embodiment provides a method for determining the target based on the known geometric scale of the target: the actual spacing of the feature points is preset during target design. The pixel spacing of the corresponding feature points in the image is measured to be Then it is acceptable. The above measurements can be completed during the initial table setup and confirmation shooting. The monitoring host will... The target ID is bound to the target point code and written to the location data table. When using a fixed camera array and the relative distance between the camera and the target point is relatively stable... It can be considered a constant and used for long-term conversion of the target point; when using UAVs for inspection and the shooting distance changes with the flight path points, the monitoring host can adjust the target based on the shooting pose information or distance estimation uploaded by the UAV. Perform dynamic updates, for example, adjust the settings based on the shooting distance during each inspection mission. Reassessment and recording of source markers reduce displacement conversion errors caused by scale drift. Furthermore, when the fixed camera experiences slight zoom or minor support movement, the monitoring host can also combine target point geometric scale observations within a time window to... Fine-tuning is performed to maintain scale consistency, but the specific update algorithm is not limited.

[0095] To improve the engineering reliability of displacement data, the image analysis module can also perform quality control on the displacement calculation results: when target point decoding fails, the number of feature points is insufficient, the sharpness score is below the threshold, or the reprojection error exceeds the limit, the monitoring host can mark the displacement frame as invalid and preferably compensate by interpolating with nearby valid frames, calling the results of other view cameras at the same target point, or triggering a drone to re-shoot within the time window; at the same time, the cause of the anomaly and the timestamp are written to the log for subsequent review and maintenance traceability. Through the above quality control mechanism, the impact of common interferences at construction sites such as rain, fog, backlight, and personnel obstruction on the continuity of monitoring can be reduced.

[0096] After obtaining the displacement values ​​of multiple monitoring target points, the monitoring host can aggregate the target point set under the same cantilevered I-beam to generate a deformation characterization value for the corresponding location of the I-beam, which is used for subsequent risk assessment. Specifically, the monitoring host can take the maximum displacement value within the same spatial unit to form... The deformation characteristics are represented by the following methods: the most unfavorable deformation at that location; the average value is taken to represent the overall trend; or a weighted sum is performed based on the target points' locations on the I-beam to represent the deflection change at a specific location. The monitoring host correlates the deformation characteristics with the force data in the same spatial dimension, using this as input for the risk assessment and graded alarm module, thereby achieving comprehensive judgment and alarm triggering based on multi-source "force-displacement" information.

[0097] Preferably, in embodiment 5, force-displacement fusion, risk assessment, and graded alarm, after completing multi-source data binding and displacement calculation, the monitoring host fuses and analyzes the force data and displacement / attitude information within the same spatial location unit to output the monitoring results of that spatial location unit. Based on this, risk assessment and graded alarms are then performed. A spatial location unit refers to the smallest associated unit in the location data table determined by "node code—target code—spatial location." The monitoring host uses this unit as the basic object for fusion and alarming, thereby avoiding misassociation of data from different I-beams or different force paths.

[0098] In terms of data fusion processing, the monitoring host preferably performs time alignment and data cleaning first. Specifically, for force data from sensor nodes, the monitoring host aggregates it into the corresponding spatial location Loc based on the node code NodeID and sorts the included timestamps; for displacement / attitude data from the image analysis module, the monitoring host aggregates it into the corresponding spatial location Loc based on the target code TargetID and removes or marks abnormal frames (such as decoding failures, low clarity, excessive reprojection errors, etc.). Subsequently, the monitoring host uses timestamps as indexes and employs nearest neighbor matching or time window aggregation to align the force data and displacement data: for example, within a given time window... Within the time window, the force sampling value is paired with the displacement value, or the force and displacement are respectively taken as statistical values ​​(such as mean, maximum, or median) within the window before fusion, thereby reducing the impact of communication jitter and frame rate inconsistency on fusion stability. Through the above processing, the monitoring host obtains the joint "force-displacement" state quantity of the same spatial location unit within time t or time window, and writes it into the monitoring result table as the basic data of the monitoring results for user terminal display and traceability.

[0099] In terms of risk assessment, the monitoring host performs normalization and weighted calculations on the force and displacement information within the same spatial unit to obtain the risk index R. The risk index in this embodiment satisfies the following relationship: ; in, The force value is real-time, preferably a representative force value within the current moment or the current time window of the spatial location unit. The representative force value can be the output value of the pressure sensor, the output value of the tension sensor, or a combination of the two. The reference stress value is preferably the stable average value after installation, the static load acceptance value, or the reference value confirmed by humans, and is bound and stored with the node code NodeID; To allow for stress thresholds, pre-set values ​​can be made based on design calculations, construction plans, or safety control requirements, and can be configured in groups according to component type or floor level. The maximum displacement of the target point within the spatial location unit is the maximum value. The monitoring host can take the maximum value of the displacement of multiple target points under the same Loc to characterize the most unfavorable deformation. The allowable displacement threshold can be preset based on control deflection, allowable deformation, or on-site management standards; The weighting coefficients are preferably configured by the monitoring host based on engineering experience or calibration results, and different weighting combinations can be used for different monitoring stages (erection, use, dismantling). By normalizing the force and displacement separately and then summing them by weight, the risk index R can simultaneously reflect the "degree of force deviation" and the "degree of deformation", thereby improving the reliability of single-source threshold alarms.

[0100] To ensure the alarms are feasible for engineering implementation, the monitoring host preferably sets up risk level classification rules, mapping the risk indicator R to at least two or three risk levels, and configuring different response strategies for different levels. For example, alert, warning, and alarm levels can be set: when R is within the first threshold range, it is an alert level to remind attention; when R exceeds the second threshold, it is a warning level to prompt inspection or enhanced monitoring; when R exceeds the third threshold or continues to exceed the second threshold for a preset duration, it is an alarm level to prompt immediate review and action. The above-mentioned thresholds can be set according to the project configuration file and are not limited to fixed values. At the same time, the monitoring host can introduce hysteresis or duration determination to suppress frequent alarms triggered by short-term noise. For example, an escalation is only triggered when R continuously exceeds the threshold for N sampling cycles or for T seconds, thereby improving alarm stability.

[0101] When the risk indicator R meets the preset alarm conditions, the monitoring host pushes an alarm message to the user terminal. The alarm message includes at least: the target ID or set of targets that triggered the alarm, the corresponding spatial location (Loc), the risk level, the trigger data, and a timestamp; wherein the trigger data preferably includes a representative force value. Reference force value Maximum displacement The system provides risk indicators (R) and related threshold information to enable on-site personnel to quickly determine the cause and priority of handling. The monitoring host simultaneously writes alarm events to the event log and generates a review task associated with the alarm event. This task drives drone re-inspection or designated fixed camera re-capture: when a fixed camera is visible in the alarm spatial location unit and the current image quality is insufficient, the monitoring host can specify a CamID to increase the frame rate and re-capture the target area; when there are obstructions, blind spots, or a need to change the viewing angle for confirmation, the monitoring host generates a drone review inspection task TaskID, plans a flight path sequence covering the target set under that Loc, and issues it for execution. After the review task is executed, the monitoring host receives the re-captured images and recalculates the displacement, then merges it with the force data to recalculate R, achieving a closed loop of "alarm—review—re-judgment"; the review results and handling suggestions can be simultaneously pushed to the user terminal, thus forming a traceable and executable risk management process on the construction site.

[0102] Preferably, in embodiment 6, the adaptive sampling control loop and disconnection buffer retransmission are used. The monitoring host and the edge gateway constitute an adaptive sampling control loop, which dynamically adjusts the multi-source data sampling and reporting strategy under different operating conditions: "normal monitoring—risk increase—review and encryption—risk decrease." This reduces communication overhead and terminal power consumption while maintaining monitoring accuracy and response speed, and improves data integrity and traceability in unstable communication scenarios. The adaptive sampling control loop uses spatial location units as the basic control object. That is, the monitoring host can issue sampling strategies for a specific spatial location (Loc) or a group of spatial location units, without requiring uniform configuration across the entire field, thus achieving resource scheduling of "key locations, key sampling."

[0103] In terms of control triggering, the monitoring host adjusts the sampling strategy based on changes in risk level and displacement change characteristics output by the risk assessment and graded alarm module. Specifically, when the monitoring host determines that the risk level of a certain spatial location unit has increased, or detects that the displacement change rate within that spatial location unit exceeds a preset threshold, the monitoring host generates an encrypted sampling command and sends it to the corresponding image acquisition terminal and sensor node terminal. The displacement change rate can be obtained by the image analysis module from the displacement difference between adjacent time points, for example, using... The system represents the displacement increment and compares it with a preset threshold. When multiple target points exist, the maximum displacement increment or the maximum rate of change within the same spatial location unit can be used as the triggering basis to cover the most unfavorable working conditions. In addition to risk level and displacement change rate, the monitoring host can also trigger encrypted sampling based on indicators such as continuous over-limit time, number of alarms, and visual missing measurement ratio to improve the response robustness under complex working conditions.

[0104] Under the encrypted sampling strategy, the monitoring host issues at least one or more of the following control commands: First, it issues a command to the fixed camera to increase the frame rate, enabling the fixed camera to acquire and upload image data containing target points at a higher frame rate within a preset time period, thereby improving the temporal resolution of displacement changes; Second, it issues an encrypted route re-inspection command to the UAV inspection unit, which may include increasing the density of route points, shortening the spacing between route points, increasing the number of frames acquired per route point, or increasing the task priority, thereby obtaining more re-inspection images from more perspectives in a shorter time; Third, it issues commands to the sensor node group to increase the sampling frequency and reporting frequency, enabling the force data to be acquired and uploaded at a higher frequency during periods of increased risk, in order to capture rapid changes in force. To ensure that the commands can be executed, the control commands issued by the monitoring host preferably carry the control object identifier (e.g., Loc, NodeID, CamID, or TaskID) and target parameters (e.g., frame rate fps, sampling frequency). Reporting frequency The parameters (or inspection point sequence density parameters) and the effective duration or failure conditions of the command enable the edge gateway, sensor node and image acquisition end to execute according to unified constraints and automatically restore the default strategy after the conditions are met.

[0105] When the risk level decreases and the data stabilizes, the monitoring host issues a downsampling command to reduce communication and energy consumption. Data stability can include: the risk indicator R continuously falling below a preset threshold for a preset time, or the displacement change rate and force change rate continuously falling below a stability threshold for a preset time. The downsampling command can restore fixed cameras to their base frame rate, drones to their routine inspection cycle or cancel encrypted re-inspection tasks, and sensor nodes to their base sampling and reporting frequencies. To avoid system jitter caused by frequent switching, the monitoring host can introduce a hysteresis mechanism or a minimum hold time during upsampling / downsampling switching. For example, encrypted sampling can be maintained for at least T seconds before downsampling is allowed, or downsampling can only be triggered after continuous stability for N cycles, thereby improving the stability of the control loop and engineering availability.

[0106] To enhance data integrity in scenarios with unstable communication, this embodiment also provides a disconnection caching and retransmission mechanism. When communication between the monitoring host and the edge gateway or sensor node is interrupted, the edge gateway and / or sensor node caches the data locally and retransmits data packets with timestamps after communication is restored, ensuring that the monitoring host can perform continuous reconstruction and traceability analysis. Specifically, the sensor node can cache the sampled data within a recent period locally in a circular buffer or file manner. The cached data packets must at least include the node code, sampling timestamp, sensor value, sequence number, and check field. The edge gateway can queue and cache data that has been received but not successfully transmitted, and record the transmission status and retransmission count. When communication is restored, the edge gateway preferably retransmits the cached data in timestamp order, carrying the original sequence number and timestamp, enabling the monitoring host to perform deduplication, out-of-order rearrangement, and packet loss detection on the retransmitted data. When the monitoring host detects discontinuous sequence numbers or gaps in timestamps, it can mark the missing test interval as "communication missing test" and provide a notification in the monitoring results and alarm analysis. Furthermore, to avoid the instantaneous bandwidth impact caused by retransmission, the edge gateway can adopt a batch retransmission or rate-limited retransmission strategy. That is, while ensuring that real-time data is uploaded first, historical data is gradually retransmitted using idle bandwidth, thereby balancing real-time monitoring and historical integrity.

[0107] Through the aforementioned adaptive sampling control loop and disconnection buffering and retransmission mechanism, this embodiment can achieve on-demand encrypted acquisition and reliable retention of multi-source monitoring data under conditions of fluctuating network conditions, visual obstruction, and rapid changes in risk status at high-rise construction sites. It also enables the monitoring host to complete risk assessment, alarm triggering, and verification closed loop based on continuous or reconstructable data sequences, thereby improving the real-time performance, reliability, and traceability of the overall monitoring system.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-rise flower basket scaffolding, comprising a flower basket tie rod assembly, a cantilevered I-beam, and a scaffolding frame, wherein the cantilevered I-beam is connected to the lower concrete wall; Its features are, The flower basket pull rod assembly includes an upper pull rod, a flower basket, and a lower pull rod. The upper pull rod and the lower pull rod are both threaded into the flower basket. The end of the upper pull rod away from the flower basket is provided with an upper connecting plate that connects to the upper concrete wall, and the end of the lower pull rod away from the flower basket is provided with a lower connecting plate that connects to the cantilevered I-beam. The upper connecting plate is connected to the upper mounting plate, and the upper mounting plate is connected to a pressure sensor that abuts against the upper concrete wall. A tension sensor is connected between the upper pull rod and the lower pull rod, and the two ends of the tension sensor are connected to the upper pull rod and the lower pull rod respectively through adjustable length connecting ropes. Each cantilevered I-beam is equipped with a readable coded marker, and each cantilevered I-beam corresponds to a unique code. It also includes a monitoring host for receiving real-time data from pressure and tension sensors and for storing the coded information of the markers.

2. The high-rise flower basket scaffolding according to claim 1, characterized in that, The upper mounting plate is provided with a vertically arranged elongated connecting hole for the screw to pass through, and the detection part of the pressure sensor is pressed against the surface of the concrete wall.

3. The high-rise flower basket scaffolding according to claim 1, characterized in that, Both ends of the upper and lower pull rods are connected to connecting rings, and quick-release buckles are connected to the connecting ropes at both ends of the tension sensor. At least one end of the connecting rope is connected to an adjusting buckle to adjust the tension of the connecting rope.

4. The high-rise flower basket scaffolding according to claim 1, characterized in that, The identifier is a QR code, barcode, or RFID tag.

5. A monitoring system for high-rise flower basket interlocking scaffolding, characterized in that, The high-rise flower basket scaffolding described in claim 1 comprises: The sensor node group is used to collect the force data at the corresponding position of the monitored cantilevered I-beam. Each sensor node in the sensor node group has at least a unique node code. A set of monitoring target points is set below the cantilevered I-beam. The monitoring target points are coded targets that can be identified by machine vision, and each monitoring target point has a unique target point code. The image acquisition subsystem includes multiple fixed cameras and / or cameras mounted on drones, used to acquire image data containing monitoring target points; An edge gateway is used to receive and forward force data from a group of sensor nodes. The monitoring host is used to receive the force data and image data, establish and maintain the position data of "node code - target code - spatial position", analyze the image data to obtain target displacement / attitude information, and fuse the force data and target displacement / attitude information to output monitoring results and alarm information. The user terminal is connected to the monitoring host and is used to display the monitoring results and alarm information.

6. The monitoring system for high-rise flower basket interlocking scaffolding according to claim 5, characterized in that, The image acquisition subsystem includes: a fixed camera array, whose installation position and field of view are configured to cover multiple monitoring target points and form overlapping fields of view to support multi-view joint solution; and / or The UAV inspection unit performs photography according to the inspection tasks issued by the monitoring host. The inspection tasks include at least the sequence of flight path points, target coverage constraints, and photography triggering conditions. The monitoring host generates the route point sequence based on location data and / or BIM data, enabling the UAV to collect image data containing at least one monitoring target point at each route point and upload it with a timestamp.

7. The monitoring system for high-rise flower basket scaffolding according to claim 5, characterized in that, The monitoring host includes an image analysis module, used for target detection, target encoding and decoding, and displacement calculation of image data, and to calculate at least the following relationship: Each monitoring target at time Displacement amount: ； in, For displacement, and The monitoring target points are respectively at time. Compared to the pixel coordinates in the image coordinate system at the initial moment. The pixel-to-actual length conversion factor is determined by the camera calibration parameters and the geometric scale of the target point; the monitoring host generates the deformation characterization quantity of the corresponding position of the cantilever I-beam based on the displacement of multiple monitoring target points.

8. The monitoring system for high-rise flower basket scaffolding according to claim 5, characterized in that, The monitoring host includes a multi-source binding module for automatically binding force data and image data in the spatial location dimension. The multi-source binding module establishes the binding relationship using at least one or a combination of the following information: The mounting pose of the fixed camera and the encoding of the identifiable target points within its field of view; The timestamp and shooting pose information of the drone during shooting, and the corresponding target point code; Preset constraints for node encoding and target encoding within the same location unit.

9. The monitoring system for high-rise flower basket interlocking scaffolding according to any one of claims 5-8, characterized in that, The monitoring host and the edge gateway form an adaptive sampling control loop: when the monitoring host determines that the risk level has increased or detects that the displacement change rate exceeds the threshold, it sends an instruction to the fixed camera to increase the frame rate and / or sends an encrypted route re-inspection instruction to the UAV inspection unit, and sends an instruction to the sensor node group to increase the sampling frequency and reporting frequency; when communication is interrupted, the edge gateway and / or sensor nodes cache the data locally and retransmit the data packets with timestamps after communication is restored.

10. The monitoring system for high-rise flower basket interlocking scaffolding according to any one of claims 5-8, characterized in that, The monitoring host includes a risk assessment and graded alarm module, used to calculate risk indicators based on force data and displacement data and trigger alarms. The risk indicators must at least meet the following requirements: ； in, This represents the real-time force value. As the reference force value, To allow for a force threshold, This refers to the maximum displacement of the monitored target point within the same spatial location unit. To allow displacement threshold, For weighting coefficients; when When preset conditions are met, the monitoring host pushes alarm information containing target code, spatial location, risk level, trigger data and timestamp to the user terminal, and automatically generates a review task to drive the drone to re-inspect or the designated fixed camera to re-shoot.

Images