The curtain wall construction hanging basket at the light well relies on the support system of the scaffolding

Abstract

The present invention relates to the technical field of building construction equipment, and discloses a support system for a curtain wall construction hanging basket at a light shaft, which relies on a scaffold for force bearing. The system includes a scaffold stress monitoring module, a hanging basket attitude solving module, a support force distribution module, a structural deformation warning module, and an adaptive leveling module. The scaffold stress monitoring module collects and processes stress data; the hanging basket attitude solving module generates dynamic balance parameters; the support force distribution module optimizes the support force distribution path; the structural deformation warning module predicts the deformation trend and outputs the warning level; the adaptive leveling module adjusts the hydraulic extension and contraction amount of the scaffold support legs to calibrate the level of the hanging basket platform. The modules work together to achieve real-time monitoring and precise control of scaffold stress and hanging basket attitude, which can effectively improve construction safety, construction accuracy and efficiency, and also has adaptive adjustment capabilities to adapt to complex construction environments and meet the needs of curtain wall construction at light shafts.

Description

Technical Field

[0001] The invention relates to the technical field of building construction equipment, in particular to a support system for a curtain wall construction hanging basket at a light well, which is supported by a scaffold. Background Art

[0002] In the field of building construction, curtain wall construction at light wells is a complex and challenging task. With the increasing diversity and sophistication of modern architectural designs, the scale and structure of light wells are becoming increasingly complex, placing higher demands on construction equipment and technology. When constructing curtain walls at light wells, workers rely on construction baskets to perform tasks such as curtain wall installation, maintenance, and cleaning. However, existing construction basket support methods present numerous problems.

[0003] Traditional construction hanging basket support methods are often relatively simple and direct, mostly by hanging the hanging basket directly on the building structure, such as the edge of the roof or pre-buried hanging points. When used in light wells, this method faces the problem of uneven structural stress. The building structures around the light well may have different load-bearing capacities and rigidity due to design or functional requirements. Directly hanging the hanging basket can easily cause some structures to be subjected to excessive stress, increasing the risk of structural damage. For example, in some areas constructed with lightweight materials at the edges of light wells, long-term exposure to the concentrated load of the hanging basket may cause wall cracking, local deformation, etc., which not only affects construction safety, but also poses a threat to the overall structural stability of the building.

[0004] In addition, existing support methods have a weak ability to control the posture of the hanging basket. During the construction process, the hanging basket will be affected by various factors, such as wind, the operation of the construction workers, and the shaking of the hanging basket itself, causing the hanging basket to become unstable, such as tilting and swinging. When the hanging basket is in an unstable state, it is difficult for construction workers to accurately perform curtain wall construction operations, reducing construction efficiency and quality. For example, when installing curtain wall glass, the shaking of the hanging basket makes it difficult to accurately align the glass, increasing the difficulty of installation and even causing accidents such as glass shattering. Moreover, the instability of the hanging basket also poses a direct threat to the safety of construction workers, increasing the possibility of safety accidents such as falling from height.

[0005] At the same time, the existing support system lacks real-time monitoring and effective early warning mechanisms for the operating status of scaffolding and hanging baskets. During the construction process, the safety status of the scaffolding and hanging baskets is crucial, but traditional methods are unable to obtain timely information on stress changes in the scaffolding nodes and the dynamic balance parameters of the hanging baskets. Once the scaffolding has structural deformation, loose connections, or abnormal swinging of the hanging basket, it is difficult to detect and take corresponding measures in a timely manner. The problem is often not discovered until it develops to a more serious level, missing the best time to deal with it, and then causing serious safety accidents. For example, a node of the scaffolding gradually loosens due to long-term stress, but due to the lack of real-time monitoring, the construction workers fail to discover it in time. As the loosening worsens, it may eventually lead to partial collapse of the scaffolding.

[0006] In addition, in actual construction, the construction environment at the light well is complex and changeable. Different construction stages, different weather conditions, and different construction process requirements all require flexible adjustment of the support system. However, most existing support systems lack adaptive adjustment capabilities, making it difficult to adjust the support force distribution, the posture of the hanging basket, and the structural state of the scaffolding in a timely manner according to actual conditions. They are unable to meet the complex and changing construction needs, and have seriously restricted the construction progress and quality. In summary, the development of a support system for curtain wall construction hanging baskets at light wells that can solve the above problems and rely on scaffolding to bear the force has important practical significance. Summary of the Invention

[0007] The purpose of the present invention is to provide a support system for a curtain wall construction hanging basket at a light well, which relies on a scaffold to bear the force, so as to solve the problems raised in the above-mentioned background technology.

[0008] To achieve the above-mentioned object, the present invention provides the following technical solution: a support system for a curtain wall construction hanging basket at a light well relying on a scaffold for bearing force, the system comprising:

[0009] The scaffolding stress monitoring module is used to collect real-time stress data of scaffolding nodes through multi-source sensors, including strain gauge arrays, inclinometers, and distributed pressure sensors, and to perform noise separation and feature extraction on the stress data using a time-frequency analysis algorithm;

[0010] The gondola posture calculation module is used to combine the laser point cloud scanning results with the sling tension data to construct the gondola kinematic constraint equations and generate the gondola dynamic balance parameters through an adaptive filtering algorithm;

[0011] A support force distribution module is used to generate an optimized support force distribution path for the scaffolding nodes through a dynamic programming algorithm based on the real-time stress data of the scaffolding nodes and the dynamic balance parameters of the hanging basket;

[0012] A structural deformation warning module is used to predict the deformation trend of the scaffolding nodes and output a warning level through a transfer learning model based on the support force optimization distribution path and the historical deformation data of the scaffolding;

[0013] The adaptive leveling module is used to adjust the hydraulic extension and contraction of the scaffolding support legs through a fuzzy control algorithm according to the warning level and the dynamic balance parameters of the hanging basket, so as to realize dynamic horizontal calibration of the hanging basket platform.

[0014] Preferably, the execution steps of the scaffolding stress monitoring module include:

[0015] Performing wavelet packet decomposition on the output signal of the strain gauge array to extract stress fluctuation components in different frequency bands, and constructing a topological network of stress transmission paths based on the spatial distribution density of the distributed pressure sensors;

[0016] Nonlinear interpolation is performed on the stress fluctuation components based on the topological network to generate a three-dimensional stress field distribution diagram of the scaffolding node.

[0017] Preferably, the execution steps of the hanging basket attitude solution module include:

[0018] Performing Euler angle transformation on the laser point cloud scanning result to generate an initial pose matrix of the hanging point of the hanging basket;

[0019] Fusing the time domain integration result of the sling tension data with the initial posture matrix to construct a kinematic constraint equation for the hanging basket;

[0020] The kinematic constraint equation is iteratively optimized by a sliding window least square method, and the dynamic balance parameters of the hanging basket are output.

[0021] Preferably, the execution steps of the support force distribution module further include:

[0022] Construct a bearing capacity threshold model for scaffolding nodes, including the yield strength of node materials, the upper limit of connector torque, and foundation settlement tolerance;

[0023] The support force optimization distribution path is input into a Monte Carlo simulator to generate a probability distribution heat map of node forces, and support forces are dynamically diverted to nodes that exceed a bearing capacity threshold according to the probability distribution heat map.

[0024] Preferably, the execution steps of the structural deformation warning module further include:

[0025] Performing time series slicing on the historical deformation data of the scaffold to construct a temporal convolution kernel of deformation features;

[0026] Performing a correlation analysis between the temporal convolution kernel and the current deformation trend through a gated recurrent unit network to generate a deformation acceleration prediction value;

[0027] Based on the deformation acceleration prediction value, warning level thresholds are divided and graded alarm signals are triggered.

[0028] Preferably, the scaffolding stress monitoring module also includes: introducing a generative adversarial network, wherein the generator is used to simulate the stress distribution under extreme load conditions, the discriminator is used to evaluate the distribution consistency between the measured stress data and the simulated data, and the parameters of the generator and the discriminator are optimized through the back propagation algorithm.

[0029] Preferably, the execution steps of the hanging basket attitude solution module further include:

[0030] Construct a dynamic model of the basket swing, including the elastic deformation coefficient of the sling, the wind load disturbance factor, and the center of mass offset of the basket;

[0031] The dynamic balance parameters of the hanging basket are input into the dynamic model, and the stability domain boundary of the hanging basket swing is solved by the Hamiltonian optimization algorithm;

[0032] The weight coefficient of the kinematic constraint equation of the hanging basket is dynamically modified according to the boundary of the stability region.

[0033] Preferably, the execution steps of the support force distribution module also include: introducing a dynamic load balancing module, predicting the spatiotemporal variation trend of the hanging basket load during the construction process through a long short-term memory network, coupling the spatiotemporal variation trend with the support force optimization distribution path for analysis, and generating an adaptive reinforcement strategy for the scaffolding node.

[0034] Preferably, the execution steps of the structural deformation warning module also include: constructing a multi-sensor data residual function, calculating the Mahalanobis distance between historical deformation data and real-time deformation data, minimizing the residual function through a particle swarm optimization algorithm, and outputting a confidence interval of the deformation prediction value.

[0035] Preferably, the execution steps of the adaptive leveling module also include: performing dead zone compensation on the adjustment process of the hydraulic expansion and contraction amount, constructing an inverse model of the hydraulic cylinder response hysteresis, and designing a control law for the inverse model through the Lyapunov stability criterion.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] To enhance safety, the scaffolding stress monitoring module collects real-time stress data from scaffolding nodes using multiple sensors, including strain gauge arrays, inclinometers, and distributed pressure sensors. It then utilizes a time-frequency analysis algorithm for noise separation and feature extraction, enabling precise monitoring of the scaffolding's stress state. A generative adversarial network (GAN) is introduced to simulate stress distribution under extreme load conditions. A discriminator evaluates the distribution consistency between measured and simulated data, optimizing both the generator and discriminator parameters. This results in more comprehensive and accurate scaffolding stress monitoring, enabling early detection of potential stress anomalies and effectively preventing scaffolding collapse accidents caused by stress concentration. The structural deformation warning module, based on optimized support force distribution paths and historical scaffolding deformation data, uses a transfer learning model to predict deformation trends at scaffolding nodes and output warning levels. Furthermore, a residual function for multi-sensor data is constructed, calculating the Mahalanobis distance between historical and real-time deformation data. A particle swarm optimization algorithm is then used to determine the confidence interval for the output deformation prediction value. This enables precise early warning of scaffolding structural deformation, significantly enhancing construction safety.

[0038] To improve construction accuracy and efficiency, the basket attitude calculation module combines laser point cloud scanning results with cable tension data to construct kinematic constraint equations for the basket. Using an adaptive filtering algorithm, the basket's dynamic balance parameters are generated, ensuring the basket maintains a stable attitude throughout the construction process. A dynamic model for basket swing is constructed to determine the stability domain boundaries for the basket's swing, and the weight coefficients of the kinematic constraint equations are dynamically modified, further improving the accuracy of the basket's attitude control. This enables construction workers to perform curtain wall construction operations more accurately from a stable platform, reducing construction errors caused by basket sway, thereby improving construction quality and efficiency. The support force distribution module uses a dynamic programming algorithm to generate an optimized support force distribution path for scaffolding nodes based on real-time stress data from scaffolding nodes and the basket's dynamic balance parameters. This optimizes support force distribution, avoids excessive stress on local nodes, and ensures overall scaffolding stability, providing a solid foundation for smooth construction. Furthermore, a dynamic load balancing module is introduced to predict the spatiotemporal variation of the basket load during construction. Coupled with the analysis of the optimized support force distribution path, an adaptive reinforcement strategy is generated, further improving the stability and reliability of the construction process.

[0039] In terms of adaptive adjustment capabilities, the adaptive leveling module uses a fuzzy control algorithm to adjust the hydraulic expansion and contraction of the scaffolding support legs based on the warning level and the dynamic balance parameters of the hanging basket, achieving dynamic horizontal calibration of the hanging basket platform. Deadband compensation is implemented during the hydraulic expansion and contraction adjustment process, an inverse model of the hydraulic cylinder response hysteresis is constructed, and the control law is designed using the Lyapunov stability criterion, ensuring the accuracy and stability of the adaptive leveling process. When faced with changes in the construction environment, such as changes in wind speed or relocation of construction personnel, the system can respond quickly, automatically adjusting the hanging basket's posture and scaffolding support force to adapt to complex and changing construction needs, reducing manual intervention and improving the level of construction automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a working principle diagram of the support system for the curtain wall construction hanging basket at the light well according to the present invention, which relies on the scaffolding for load bearing;

[0041] Figure 2 Flowchart for extracting steady-state physiological features for the adaptive filtering algorithm;

[0042] Figure 3 This is the workflow diagram of the multi-dimensional decision model-feature fusion layer;

[0043] Figure 4 This is a flow chart for the hierarchical processing of scaffolding deformation warning. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0045] See also Figures 1-4 The present invention provides a support system for a curtain wall construction hanging basket at a light well, which relies on a scaffold to bear the force. The specific implementation method thereof is described in detail below.

[0046] Scaffolding Stress Monitoring Module: This module collects real-time stress data from scaffolding nodes using multiple sensors, including strain gauge arrays, inclinometers, and distributed pressure sensors. These sensors are strategically placed at key scaffolding nodes to accurately capture stress information. After collecting the data, a time-frequency analysis algorithm is used to perform noise separation and feature extraction on the stress data, removing interfering signals and highlighting effective stress characteristics, providing a reliable data foundation for subsequent analysis.

[0047] The hanging basket posture calculation module combines laser point cloud scanning results with cable tension data to construct the kinematic constraint equations for the hanging basket and generates the dynamic balance parameters of the hanging basket using an adaptive filtering algorithm. The multi-dimensional posture data, laser point cloud scanning results, and cable tension data reflect the status of the hanging basket from different perspectives. By fusing these data with the adaptive filtering algorithm, the parameters required for dynamic balance are accurately calculated to maintain the hanging basket's stability during construction.

[0048] The Support Force Distribution Module uses a dynamic programming algorithm to generate an optimized support force distribution path for each scaffolding node based on real-time stress data from the scaffolding node and the dynamic balance parameters of the hanging basket. This algorithm comprehensively considers the load-bearing capacity of each node, the current stress state, and the balance requirements of the hanging basket to calculate the most reasonable support force distribution plan, ensuring that the force applied to each scaffolding node is evenly distributed and within a safe range.

[0049] Structural Deformation Warning Module: Based on the optimized support force distribution path and historical scaffolding deformation data, a transfer learning model is used to predict the deformation trend of scaffolding nodes and output a warning level. The transfer learning model draws on deformation data and experience from previous similar scaffolding structures, combined with the current support force distribution, to accurately predict node deformation trends. Different warning levels are assigned based on the predicted results, allowing timely implementation of appropriate measures to ensure construction safety.

[0050] Adaptive Leveling Module: Based on the warning level and the dynamic balance parameters of the hanging basket, a fuzzy control algorithm adjusts the hydraulic expansion and contraction of the scaffolding support legs to achieve dynamic horizontal calibration of the hanging basket platform. The fuzzy control algorithm intelligently controls the hydraulic expansion and contraction based on the warning level and the dynamic balance parameters of the hanging basket, ensuring that the hanging basket platform remains level, providing a stable working platform for construction workers.

[0051] The present invention will be further described below in conjunction with Examples 1 to 5:

[0052] Example 1

[0053] This embodiment mainly involves more detailed execution steps of the scaffolding stress monitoring module.

[0054] In actual construction, the strain gauge array is precisely attached to the key stress-bearing parts of the scaffolding node. Its output signal contains rich stress information, but it is inevitably mixed with various noises. At this time, the output signal of the strain gauge array is subjected to wavelet packet decomposition, which can decompose the signal into different frequency bands. Assume that the output signal of the strain gauge array is After wavelet packet decomposition, the stress fluctuation components in different frequency bands are obtained ( , is the number of frequency bands obtained by decomposition).

[0055] A topological network of stress transmission paths is constructed based on the spatial density of the distributed pressure sensors. Distributed pressure sensors are located at different locations within the scaffold, and their density reflects the importance of stress transmission in different areas. By analyzing the location and measurement data of these sensors, the stress transmission paths within the scaffold structure are determined and a topological network is constructed. This network clearly illustrates the stress propagation relationships between the various nodes of the scaffold.

[0056] Based on the topological network, nonlinear interpolation is performed on the stress fluctuation components. Based on the constructed topological network, a nonlinear interpolation method is used to obtain a more continuous and accurate stress distribution, as the stress fluctuation components have a certain degree of spatial discreteness. This method generates a three-dimensional stress field distribution map for the scaffolding nodes. This map displays the stress magnitude and distribution of each scaffolding node in an intuitive three-dimensional format, allowing construction personnel and management personnel to understand the scaffolding stress state in real time and promptly identify potential stress concentration areas, providing strong support for ensuring the safety and stability of the scaffolding.

[0057] At the same time, the scaffolding stress monitoring module also introduces a generative adversarial network. The generator is used to simulate the stress distribution under extreme load conditions. In actual construction, although the scaffolding is in a normal load state most of the time, it may also encounter extreme conditions such as strong winds and sudden drops of heavy objects. The generator simulates the stress distribution under these extreme load conditions and assumes that the simulated stress distribution data is The discriminator is used to evaluate the measured stress data With simulated data The consistency of the distribution is measured by calculating the difference between the two, such as using metrics such as the Kullback-Leibler Divergence. The backpropagation algorithm is used to optimize the parameters of the generator and discriminator, enabling the generator to more accurately simulate stress distribution under extreme loads, and the discriminator to more precisely assess the difference between simulated and measured data. This improves the accuracy and reliability of scaffolding stress monitoring and identifies potential safety hazards in advance.

[0058] Example 2

[0059] This embodiment focuses on the more in-depth execution steps of the basket posture calculation module.

[0060] After obtaining the laser point cloud scanning results, the laser point cloud scanning results are subjected to Euler angle transformation. Laser point cloud scanning can obtain spatial information around the hanging point of the hanging basket, but this information needs to be converted into a pose matrix form that is easy to analyze. Let the laser point cloud scanning result be represented as a point set , after Euler angle transformation, the initial pose matrix of the hanging point of the hanging basket is generated The Euler angle transformation converts the point cloud data into a suitable coordinate system by rotating around different coordinate axes, thereby determining the initial position and posture of the hanging point of the gondola.

[0061] The time domain integration result of the cable tension data is integrated with the initial posture matrix to construct the kinematic constraint equation of the hanging basket. The cable tension data changes with time, and the time domain integration is performed on it. Let the cable tension be , and its time domain integration result is . Combine this integral result with the initial pose matrix Combined with the motion law and mechanical characteristics of the hanging basket, the kinematic constraint equation of the hanging basket is constructed. This equation describes the relationship between the various parameters of the hanging basket during its movement, providing an important basis for the subsequent accurate calculation of the dynamic balance parameters of the hanging basket.

[0062] The kinematic constraint equations are iteratively optimized using the sliding window least squares method to output the dynamic balance parameters of the hanging basket. The sliding window least squares method is an optimization method commonly used in time series data processing. In the present embodiment, as time passes, new measurement data are constantly generated. The sliding window least squares method performs a least squares fit on the data within a time window of a fixed length, continuously adjusting the parameters in the equation so that the kinematic constraint equations can more accurately reflect the actual motion state of the hanging basket. After multiple iterative optimizations, accurate dynamic balance parameters of the hanging basket, such as the horizontal offset and tilt angle of the hanging basket, are finally output. These parameters are crucial for ensuring the stability of the hanging basket during construction.

[0063] In addition, a dynamic model of the hanging basket swing is constructed, which includes the elastic deformation coefficient of the sling , wind load disturbance factor and the center of mass offset of the basket . Elastic deformation coefficient of sling It reflects the elastic deformation characteristics of the sling under stress, and the wind load disturbance factor It reflects the degree of influence of wind on the swing of the hanging basket and the offset of the center of mass of the hanging basket. It describes the deviation of the center of mass of the basket from the ideal position.

[0064] The dynamic balance parameters of the hanging basket are input into the dynamic model, and the Hamiltonian optimization algorithm is used to determine the stability region boundary of the hanging basket's swing. The Hamiltonian optimization algorithm is an energy-based optimization method. In this embodiment, the Hamiltonian optimization algorithm analyzes energy changes during the hanging basket's swing and, in combination with the hanging basket's dynamic balance parameters, calculates the stability region boundary of the hanging basket's swing. This stability region boundary determines the range within which the hanging basket can swing stably under different conditions.

[0065] The weight coefficients of the kinematic constraint equations for the hanging basket are dynamically modified based on the stability domain boundary. Once the stability domain boundary is determined, the weight coefficients of the kinematic constraint equations for the hanging basket are adjusted based on the boundary conditions. For example, if the hanging basket is found to be close to the stability domain boundary, the weights of stability-related parameters are appropriately increased, making the kinematic constraint equations more focused on maintaining the stability of the hanging basket, thereby further improving the accuracy and stability of the hanging basket posture solution and ensuring construction safety.

[0066] Example 3

[0067] This embodiment describes in detail the further execution steps of the support force distribution module.

[0068] First, the bearing capacity threshold model of the scaffolding node is constructed, which includes the yield strength of the node material , Upper limit of torque of connecting parts and foundation settlement tolerance Node material yield strength It refers to the stress value when the node material begins to undergo plastic deformation. It is one of the important indicators to measure the bearing capacity of the node. It determines the maximum torque that the connector can withstand while ensuring a secure connection. If this limit is exceeded, the connector may become loose or even fail. The maximum allowable settlement of the foundation is stipulated. Once the foundation settlement exceeds this tolerance, the overall structure of the scaffolding may become unstable.

[0069] The optimized support force distribution path is input into a Monte Carlo simulator to generate a heat map of the probability distribution of node forces. A Monte Carlo simulator is a simulation tool based on probability statistics. In this embodiment, the optimized support force distribution path is used as input, and a large number of random simulation calculations are performed to obtain the probability distribution of forces on each node under different circumstances. These probability distributions are displayed as heat maps, with darker colors indicating a greater probability that the force on the node exceeds a certain threshold. Construction personnel and management personnel can intuitively identify which nodes are at greater risk of force exposure from the heat map.

[0070] According to the probability distribution heat map, the nodes that exceed the bearing capacity threshold are dynamically diverted. When it is found from the heat map that the force on some nodes exceeds the bearing capacity threshold, timely measures are taken to dynamically divert the force. For example, for a certain node, its actual force Exceeds the bearing capacity threshold determined by the yield strength of the node material, the upper limit of the torque of the connection and the foundation settlement tolerance By adjusting the support force distribution of the surrounding nodes, part of the force is transferred to the nodes with surplus bearing capacity, ensuring that the force on each node is within a safe range and ensuring the overall stability of the scaffolding.

[0071] At the same time, the support force distribution module introduces a dynamic load balancing module to predict the temporal and spatial variation trend of the hanging basket load during the construction process through the long short-term memory network. The long short-term memory network (LSTM) is a special recurrent neural network that can effectively handle the long-term dependency problem in time series data. In this embodiment, the LSTM network predicts the variation trend of the hanging basket load at different time and spatial positions during the construction process by learning historical hanging basket load data and related information such as construction time and location. Let the predicted load variation trend be ,in Indicates time, Indicates spatial location.

[0072] The spatiotemporal variation trend Coupled with the optimized support force distribution path, this analysis generates adaptive reinforcement strategies for scaffolding nodes. By comparing the predicted load trends with the current optimized support force distribution path, we identify nodes likely to bear significant loads in the future. For these nodes, we develop adaptive reinforcement strategies, such as adding support structures or replacing stronger connectors. These preventative measures ensure the scaffolding can safely and stably withstand the varying basket loads during construction.

[0073] Example 4

[0074] This embodiment focuses on describing the more detailed execution steps of the structural deformation warning module.

[0075] The historical deformation data of the scaffold is sliced in time series to construct a time convolution kernel of the deformation feature. The historical deformation data of the scaffold is accumulated over time and reflects the deformation of the scaffold at different stages. Through time series slicing, the historical deformation data is divided according to a certain time interval, and the time interval is set to , obtaining a series of time-segment data. Then, a temporal convolution kernel for deformation features is constructed based on these time-segment data. The temporal convolution kernel can extract temporal features of deformation data, such as deformation trends and frequency of change.

[0076] The temporal convolution kernel and the current deformation trend are analyzed for correlation through the gated recurrent unit network to generate a deformation acceleration prediction value. The gated recurrent unit network (GRU) is a neural network that can effectively handle the information transmission and forgetting problems in time series data. In this embodiment, the GRU network combines the historical deformation features contained in the temporal convolution kernel with the current deformation trend to perform an in-depth correlation analysis. Through this analysis, the deformation acceleration of the scaffolding node is predicted. Let the predicted deformation acceleration be The deformation acceleration reflects the change of the node deformation speed and is of great significance for judging the stability of the scaffolding.

[0077] Based on the deformation acceleration prediction value, the warning level threshold is divided and the graded alarm signal is triggered. According to actual engineering experience and safety standards, different deformation acceleration thresholds are set to divide the warning level. For example, when the deformation acceleration Less than the first threshold When Greater than or equal to and is less than the second threshold A yellow warning is issued when Greater than or equal to Once the predicted deformation acceleration reaches the corresponding threshold, a graded alarm signal is immediately triggered, reminding construction workers and management personnel to take appropriate measures, such as stopping construction and inspecting and reinforcing scaffolding, to ensure construction safety.

[0078] In addition, a multi-sensor data residual function is constructed to calculate the Mahalanobis distance between historical deformation data and real-time deformation data. Assume that the historical deformation data is , the real-time deformation data is , the residual function of multi-sensor data is , the Mahalanobis distance is The Mahalanobis distance can take into account the covariance information of the data and more accurately measure the degree of difference between two sets of data. By calculating the Mahalanobis distance, the deviation between real-time deformation data and historical data can be determined.

[0079] The residual function is minimized using a particle swarm optimization algorithm, which outputs a confidence interval for the deformation prediction value. Particle Swarm Optimization (PSO) is an optimization algorithm based on swarm intelligence. In this embodiment, the PSO algorithm continuously adjusts the positions and velocities of particles to find the parameter combination that minimizes the residual function. After multiple iterations of optimization, a confidence interval for the deformation prediction value is obtained. This confidence interval reflects the reliability of the deformation prediction value. Construction personnel and management personnel can use this confidence interval to more accurately assess scaffolding deformation and formulate more appropriate safety measures.

[0080] Example 5

[0081] This embodiment focuses on the in-depth execution steps of the adaptive leveling module. In the adaptive leveling module, the hydraulic expansion and contraction of the scaffolding support legs are adjusted according to the warning level and the dynamic balance parameters of the hanging basket. In this process, the dead zone compensation of the hydraulic expansion and contraction adjustment is crucial.

[0082] When the hydraulic system is running, there is a dead zone phenomenon due to factors such as friction between mechanical components and the viscosity of hydraulic oil. When the input signal changes within a specific range, the hydraulic cylinder will not respond immediately, which will affect the leveling accuracy of the hanging platform. To solve this problem, an inverse model of the hydraulic cylinder response hysteresis is constructed. Assume that the control signal input to the hydraulic cylinder is , the hydraulic expansion and contraction output of the hydraulic cylinder is In practice, the response of the hydraulic cylinder can be expressed as a complex nonlinear function , but there is a dead zone effect. The inverse model constructed The purpose is to adjust the hydraulic expansion and contraction of the output Perform reverse deduction to obtain the input control signal that can eliminate the dead zone effect For example, within the dead zone, the forward model may have input signal changes but no output response, while the inverse model can calculate the appropriate input signal based on the expected output to compensate for the deviation caused by the dead zone.

[0083] The control law of the inverse model is designed by Lyapunov stability criterion to ensure the stability of the entire system. Lyapunov stability criterion is a powerful tool for analyzing the stability of dynamic systems. The Lyapunov function is defined as , It is about the system state variable (here the hydraulic expansion and contraction ) and satisfies ( ), .right Ask about time The derivative of When designing the control law, it is necessary to use This means that as time goes by, the value of the Lyapunov function decreases and the energy of the system gradually decreases, thus ensuring the stability of the system.

[0084] In actual operation, the safety status of the scaffolding and hanging basket is first determined based on the warning level. If the warning level indicates a certain risk, the hydraulic expansion and contraction amount that needs to be adjusted is determined based on the dynamic balance parameters of the hanging basket. The input control signal after considering dead zone compensation is calculated through the inverse model. , and then the control law designed according to Lyapunov stability criterion is used to control the input signal Adjustments and optimizations are made. For example, if the basket is detected to be tilted to one side, the degree of tilt (provided by the basket's dynamic balance parameters) determines whether the hydraulic extension and retraction of certain support legs needs to be increased or decreased. An inverse model is used to calculate a preliminary control signal, which is then fine-tuned using the Lyapunov stability criterion to ensure stable hydraulic system operation during the adjustment process, without oscillation or loss of control. This precise control allows for high-precision dynamic level calibration of the scaffolding support legs, creating a safe and stable working environment for curtain wall construction workers and improving construction efficiency and quality.

[0085] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0086] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A support system for curtain wall construction hanging basket at light well relying on scaffolding, characterized in that: include: The scaffolding stress monitoring module is used to collect real-time stress data of scaffolding nodes through multi-source sensors, including strain gauge arrays, inclinometers, and distributed pressure sensors, and to perform noise separation and feature extraction on the stress data using a time-frequency analysis algorithm; The gondola posture calculation module is used to combine the laser point cloud scanning results with the sling tension data to construct the gondola kinematic constraint equations and generate the gondola dynamic balance parameters through an adaptive filtering algorithm; A support force distribution module is used to generate an optimized support force distribution path for the scaffolding nodes through a dynamic programming algorithm based on the real-time stress data of the scaffolding nodes and the dynamic balance parameters of the hanging basket; A structural deformation warning module is used to predict the deformation trend of the scaffolding nodes and output a warning level through a transfer learning model based on the support force optimization distribution path and the historical deformation data of the scaffolding; The adaptive leveling module is used to adjust the hydraulic extension and contraction of the scaffolding support legs through a fuzzy control algorithm according to the warning level and the dynamic balance parameters of the hanging basket, so as to realize dynamic horizontal calibration of the hanging basket platform.

2. The support system according to claim 1, wherein: The execution steps of the scaffolding stress monitoring module include: Performing wavelet packet decomposition on the output signal of the strain gauge array to extract stress fluctuation components in different frequency bands, and constructing a topological network of stress transmission paths based on the spatial distribution density of the distributed pressure sensors; Nonlinear interpolation is performed on the stress fluctuation components based on the topological network to generate a three-dimensional stress field distribution diagram of the scaffolding node.

3. The support system according to claim 1, wherein: The execution steps of the hanging basket attitude solution module include: Performing Euler angle transformation on the laser point cloud scanning result to generate an initial pose matrix of the hanging point of the hanging basket; Fusing the time domain integration result of the sling tension data with the initial posture matrix to construct a kinematic constraint equation for the hanging basket; The kinematic constraint equation is iteratively optimized by a sliding window least square method, and the dynamic balance parameters of the hanging basket are output.

4. The support system according to claim 1, wherein: The execution steps of the support force distribution module also include: Construct a bearing capacity threshold model for scaffolding nodes, including the yield strength of node materials, the upper limit of connector torque, and foundation settlement tolerance; The support force optimization distribution path is input into a Monte Carlo simulator to generate a probability distribution heat map of node forces, and support forces are dynamically diverted to nodes that exceed a bearing capacity threshold according to the probability distribution heat map.

5. The support system according to claim 1, wherein: The execution steps of the structural deformation warning module also include: Performing time series slicing on the historical deformation data of the scaffold to construct a temporal convolution kernel of deformation features; Performing a correlation analysis between the temporal convolution kernel and the current deformation trend through a gated recurrent unit network to generate a deformation acceleration prediction value; Based on the deformation acceleration prediction value, warning level thresholds are divided and graded alarm signals are triggered.

6. The support system according to claim 2, wherein: The scaffolding stress monitoring module also includes: introducing a generative adversarial network, in which the generator is used to simulate the stress distribution under extreme load conditions, the discriminator is used to evaluate the distribution consistency between the measured stress data and the simulated data, and the parameters of the generator and the discriminator are optimized through the back propagation algorithm.

7. The support system according to claim 3, wherein: The execution steps of the hanging basket attitude solution module also include: Construct a dynamic model of the basket swing, including the elastic deformation coefficient of the sling, the wind load disturbance factor, and the center of mass offset of the basket; The dynamic balance parameters of the hanging basket are input into the dynamic model, and the stability domain boundary of the hanging basket swing is solved by the Hamiltonian optimization algorithm; The weight coefficient of the kinematic constraint equation of the hanging basket is dynamically modified according to the boundary of the stability region.

8. The support system according to claim 4, wherein: The execution steps of the support force distribution module also include: introducing a dynamic load balancing module, predicting the spatiotemporal variation trend of the hanging basket load during the construction process through a long short-term memory network, coupling the spatiotemporal variation trend with the support force optimization distribution path for analysis, and generating an adaptive reinforcement strategy for the scaffolding node.

9. The support system according to claim 5, wherein: The execution steps of the structural deformation warning module also include: constructing a multi-sensor data residual function, calculating the Mahalanobis distance between historical deformation data and real-time deformation data, minimizing the residual function through a particle swarm optimization algorithm, and outputting a confidence interval of the deformation prediction value.

10. The support system according to claim 1, wherein: The execution steps of the adaptive leveling module also include: performing dead zone compensation on the adjustment process of the hydraulic expansion and contraction amount, constructing an inverse model of the hydraulic cylinder response hysteresis, and designing a control law for the inverse model using the Lyapunov stability criterion.

Images