Control method of scaffold support system of distributed nodes and related device

Abstract

The application relates to a control method and related device of a scaffold support system of a distributed node, and relates to the field of engineering structures. The method comprises the following steps: converting the scaffold support system into a graph structure, marking nodes with average force reaching or exceeding an initial stress threshold as candidate main path nodes, marking candidate main path nodes with topological connectivity meeting preset requirements as main load-bearing path nodes, and marking remaining nodes in the graph structure as secondary load-bearing path nodes; for a first intelligent node marked as a main load-bearing path node, controlling a piezoelectric stack actuator in the first intelligent node to generate a positive displacement, increasing the normal pressure of a node connecting surface to improve the node stiffness, re-acquiring stress data of each intelligent node to calculate a stress unevenness coefficient, and stopping iterative adjustment when the stress unevenness coefficient is less than or equal to a preset convergence threshold. The above technical scheme can reasonably regulate and control the node stiffness of the scaffold support system.

Description

Technical Field

[0001] This application relates to the field of engineering structures, and in particular to a control method and related apparatus for a distributed node scaffolding support system. Background Technology

[0002] In the construction industry, scaffolding support systems are critical infrastructure for ensuring construction safety and smooth progress. With the continuous development of the construction industry, the performance and safety requirements for scaffolding support systems are also increasing. A stable and reliable scaffolding support system can provide a safe working environment for construction workers, ensure efficient construction processes, reduce the occurrence of safety accidents, and is of great significance for improving building quality and construction efficiency.

[0003] Traditional scaffolding support systems typically employ fixed structural designs and experience-based layouts when dealing with complex stress conditions. Construction workers determine parameters such as scaffold member specifications, spacing, and connection methods based on past experience. Monitoring the scaffolding's stress often relies on periodic manual inspections, observing external deformation and member connections to assess its stress state. Adjustments to scaffolding stiffness are generally made manually by adjusting the tightness of members or adding supporting components.

[0004] However, these traditional methods have significant drawbacks. Manual monitoring and adjustment methods struggle to obtain real-time and accurate stress data for each node of the scaffolding, making it impossible to promptly identify potential safety hazards. Furthermore, fixed structure designs are ill-suited to the complex and varied load conditions under different construction scenarios, easily leading to uneven stress distribution across the scaffolding nodes, increasing the risk of structural instability, and failing to effectively guarantee safety during construction. Summary of the Invention

[0005] This application provides a control method and related device for a distributed node scaffolding support system, which can reasonably adjust the node stiffness of the scaffolding support system and realize the reasonable distribution of force on the scaffolding support system.

[0006] In a first aspect, this application provides a control method for a distributed node scaffolding support system, the scaffolding support system including multiple intelligent nodes, the method comprising: After the scaffolding support system is erected and bears the load, pressure sensors distributed on each intelligent node continuously collect force data. The sliding window algorithm is used to extract features from the force data collected by each intelligent node to obtain feature data. The feature data includes the average force, peak force and rate of change of force for each window. The feature data of each intelligent node are collected, and the scaffolding support system is converted into a graph structure with intelligent nodes as vertices and member connection relationships as edges. The global mean and standard deviation of the average force of all intelligent nodes are calculated, an initial stress threshold is set, and nodes whose average force reaches or exceeds the initial stress threshold are marked as candidate main path nodes. The topological connectivity is verified by calculating the proportion of the maximum connected component of the candidate main path nodes in the graph structure. The candidate main path nodes whose topological connectivity meets the preset requirements are marked as main load-bearing path nodes, and the remaining nodes in the graph structure are marked as secondary load-bearing path nodes. For the first smart node marked as the primary load-bearing path node, the piezoelectric stack actuator inside the first smart node is controlled to generate positive displacement, increasing the positive pressure on the node connection surface to improve node stiffness. For the second smart node marked as the secondary load-bearing path node, the piezoelectric stack actuator inside the second smart node is controlled to generate negative displacement or maintain zero displacement, reducing the positive pressure on the node connection surface to reduce node stiffness. After each stiffness adjustment is completed, the stress data of each smart node is re-acquired to calculate the stress non-uniformity coefficient. When the stress non-uniformity coefficient is less than or equal to the preset convergence threshold, the iterative adjustment is stopped. The stress non-uniformity coefficient is used to characterize the coefficient of variation of the stress values ​​of all smart nodes.

[0007] By adopting the above technical solution, and transmitting feature values ​​instead of all sampled data, the communication load and storage pressure in the distributed network are significantly reduced, providing a feasible basis for real-time collaborative control of large-scale scaffolding systems. The scaffolding structure is abstracted as a graph structure, considering both the stress threshold of nodes and verifying whether candidate nodes form a continuous force-transmitting skeleton in space through the proportion of the largest connected components. This effectively avoids the "misjudgment of isolated high-stress points" problem caused by relying solely on the magnitude of stress in existing technologies, ensuring that the identified main paths truly reflect the force flow transmission paths of the structure. Mimicking the adaptive growth mechanism of bones, the stiffness of nodes on the main load-bearing paths is enhanced (simulating bone growth), while the stiffness of nodes on secondary load-bearing paths is weakened (simulating bone thinning). Micrometer-level control of node stiffness is achieved by precisely adjusting the normal pressure on the node connection surfaces using piezoelectric stack actuators. The force flow distribution of the structure is actively reconstructed, guiding the load to concentrate on the main paths, achieving optimal load-bearing efficiency with minimal material consumption.

[0008] In some embodiments, the step of using a sliding window algorithm to extract features from the force data collected by each of the intelligent nodes to obtain feature data includes: The number of sampling points for each window is calculated using the following formula: ; in, Nw Indicates the number of sampling points. L Indicates the window length. f s Indicates the sampling frequency; The average force for each window is calculated using the following formula: ; in, Indicates average force. F j Indicates the first in the window j The force values ​​at each sampling point; The peak force for each window is calculated using the following formula: ; in, F max Indicates peak force. F 1 The force value at the starting sampling point of the window. F Nw The force value at the end of the window sampling point. max Indicates the maximum value; The rate of change of force for each window is calculated using the following formula: ; in, r It represents the rate of change of force.

[0009] By employing the above technical solution, the sliding window algorithm compresses continuous raw sampling data into only three feature values ​​(average force, peak force, and rate of change) for each window. This avoids the transmission of massive amounts of raw data between distributed nodes, significantly reducing the bandwidth consumption of the communication network and enabling real-time collaborative control of large-scale scaffolding systems. The combination of the three feature values ​​characterizes the stress state of the nodes from different dimensions, completely preserving all the key information required for stress path identification with a minimal amount of data. This avoids the loss of transient information caused by simply storing average values, providing high-quality data input for the accurate identification of subsequent main load-bearing paths.

[0010] In some embodiments, controlling the piezoelectric stack actuator inside the first smart node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Calculate the control voltage, and use the control voltage to control the piezoelectric stack actuator to generate a positive displacement, thereby increasing the positive pressure on the connecting surface; The control voltage is calculated using the following formula: ; in, e ( t )= Ftarget - F cur , e ( t )express t The force error at any moment, F target Indicates the target force value. F cur express t Force value at all times u ( t )express t Control voltage at time, Δ t Indicates the control cycle. K p , K i , K d These represent the proportional, integral, and differential coefficients, respectively. e ( k ) indicates the first k Force error at each sampling time, k =0, 1, 2, ... t .

[0011] By employing the above technical solution, the proportional, integral, and derivative coefficients of the PID controller can be adjusted according to the structural characteristics of the node's location. For nodes along the primary load-bearing path, a higher proportional gain can be set to achieve rapid response; for nodes along secondary load-bearing paths, a higher derivative gain can be set to avoid oscillations during the adjustment process. This parameter adjustability allows the same control algorithm to adapt to intelligent nodes with different stiffness characteristics and load conditions, exhibiting good versatility and adaptability. The PID control process is completed locally on the microprocessor of each intelligent node. Each node independently calculates the control voltage based on feedback from its own pressure sensor, without relying on a central controller for centralized calculation. This distributed control architecture enables each node to have independent decision-making and execution capabilities. The failure of a single node does not affect the normal adjustment of other nodes, significantly improving the robustness and reliability of the system, while avoiding the communication delays and computational bottlenecks caused by a central controller.

[0012] In some embodiments, controlling the piezoelectric stack actuator inside the first smart node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Based on the topological connection relationship of the scaffolding support system, the entire topological structure is divided into multiple connected components, and the intelligent nodes in each connected component are physically connected to each other. For each connected component, an influence factor is calculated based on the force weight of each smart node and the number of connecting members in the structural topology. The force weight is the proportion of the node's force value to the total force of the structure, and the number of connecting members represents the importance of the node in the structure. The connected components are prioritized according to the order of their influence factors from largest to smallest. Based on the priority order, stiffness adjustment is performed on each connected component in turn, starting from the connected component with the largest influence factor. For each of the intelligent nodes within the same connected component, a local synchronization signal is generated through a distributed communication network to simultaneously perform stiffness adjustment within the same time window. During the adjustment process, the stiffness adjustment amount of each of the intelligent nodes is multiplied by a dynamic damping factor, which gradually increases over time. During stiffness adjustment, the rate of force change of each intelligent node and the amplitude of force fluctuation between adjacent intelligent nodes are monitored in real time. If the rate of change of force exceeds the preset safety rate threshold, or the amplitude of force fluctuation at adjacent nodes exceeds the preset fluctuation threshold, the current adjustment is paused, the control voltage is restored to the state before adjustment, and the adjustment is re-executed with a reduced adjustment step size after the stress wave dissipates.

[0013] By employing the above technical solution, the structure is divided into connected components, and the influence factor is calculated by comprehensively considering the force weight of nodes and the number of connecting members, thus achieving quantitative identification of key areas in the structure. Each connected component is adjusted sequentially according to the influence factor from largest to smallest, ensuring that areas bearing the main load transmission function complete stiffness optimization first, avoiding ineffective adjustments to non-critical areas, and significantly shortening the overall convergence time. Nodes within the same connected component simultaneously perform stiffness adjustments through local synchronization signals, avoiding mutual restraint and coupled oscillations caused by differences in adjustment timing among nodes within the component. A dynamic damping factor that gradually increases over time is introduced, making the initial adjustment relatively gentle to avoid starting shock, and gradually increasing in strength later to achieve precise adjustment, balancing the stability and speed of the adjustment process, and solving the core technical challenge in multi-actuator collaborative control.

[0014] In some embodiments, the re-collection of the stress data of each of the smart nodes to calculate the stress non-uniformity coefficient includes: The stress non-uniformity coefficient is calculated using the following formula: ; ; in, This represents the stress non-uniformity coefficient, and m represents the total number of smart nodes. Let μ be the average force value of the i-th node, and μ be the global mean of the average force of all smart nodes.

[0015] By employing the aforementioned technical solution, the coefficient of variation (the ratio of standard deviation to mean) of the stress values ​​at all intelligent nodes is calculated as the stress non-uniformity coefficient. This dimensionless index is unaffected by structural scale or the absolute value of the load, and can objectively and uniformly reflect the dispersion of stress distribution throughout the entire scaffolding support system. In biological skeletons, the dynamic balance between osteoblasts and osteoclasts ultimately strengthens the bone in areas of high stress and weakens it in areas of low stress, thus making the overall stress distribution tend towards equilibrium. Using the stress non-uniformity coefficient as the optimization target, the iterative adjustment process of the scaffolding structure accurately maps to the adaptive growth mechanism of the bone, realizing the biomimetic principle of "stress-guided structural optimization."

[0016] In some embodiments, verifying topological connectivity by calculating the percentage of the largest connected component in the graph structure of the candidate major path nodes includes: Extract the induced subgraph from the graph structure, which is composed of the candidate main path nodes and the connecting edges between the candidate main path nodes. Use a graph search algorithm to decompose the induced subgraph into connected components, identify all connected components contained in the induced subgraph, and record the node information contained in each connected component. From all connected components, the connected component with the most nodes is selected as the maximum connected component. The relative proportion of the number of nodes contained in the maximum connected component to the total number of candidate main path nodes is calculated, and the relative proportion is compared with the preset connectivity requirements. If the relative proportions meet the preset connectivity requirements, then the candidate main path nodes are determined to satisfy the topological connectivity conditions, and the current classification result is accepted. If the relative proportion does not meet the preset connectivity requirements, the candidate main path node is determined to not meet the connectivity conditions. The initial force threshold is iteratively adjusted within the force threshold range using a binary search method until the connectivity index meets the preset requirements or reaches the maximum number of iterations.

[0017] By employing the aforementioned technical solution, the induced subgraph formed by candidate main nodes is extracted and its connected components are decomposed. This requires candidate nodes to form a continuous force transmission skeleton in space, ensuring that the identified main paths can truly transmit loads. This fundamentally solves the problem of misjudging "high-stress but isolated" nodes. By calculating the relative proportion of nodes included in the largest connected component to the total number of candidate main nodes, the abstract concept of "topological connectivity" is transformed into a quantifiable numerical indicator. This indicator intuitively reflects the degree to which candidate main nodes cluster in space to form a continuous skeleton: a ratio close to 1 indicates that candidate nodes are highly connected and can form a complete force transmission path; a ratio significantly less than 1 indicates that candidate nodes are scattered and isolated, with broken force flow paths. This quantification method provides an objective and repeatable basis for connectivity judgment. When connectivity does not meet the preset requirements, the connectivity judgment result is used as a feedback signal, and a bisection method is used to iteratively adjust the initial stress threshold within the stress threshold range. If candidate nodes are too scattered (low connectivity), the threshold is increased to filter out truly high-stress nodes; if the total number of candidate nodes is too small, resulting in insufficient connectivity, the threshold is appropriately decreased to expand the candidate range. This feedback-based adaptive adjustment mechanism enables the identification parameters to automatically adapt to the mechanical properties of different structures, eliminating the need for repeated manual trial and error.

[0018] In some embodiments, converting the scaffolding support system into a graph structure with smart nodes as vertices and member connection relationships as edges includes: Based on the design drawings or actual assembly information of the scaffolding support system, each smart node is mapped to a vertex in the graph structure, and the connection relationship between adjacent smart nodes is mapped to an edge in the graph structure. An initial undirected graph reflecting the physical connection relationship of the structure is constructed, and three-dimensional coordinate information in space is associated with each vertex. Based on the force characteristic data collected by each intelligent node, a set of mechanical attributes is bound to each vertex in the initial undirected graph. The set of mechanical attributes includes the average force value, peak force value, and force change rate corresponding to the intelligent node. The static weight is determined by the sum of the average force values ​​of the two vertices connected to each edge in the initial undirected graph, and the dynamic weight is determined by the difference in the rate of change of the force of the two vertices connected to each edge. An initial weight is assigned to each edge in the initial undirected graph based on the static weight and the dynamic weight. The graph structure is constructed based on the set of vertices, the set of edges, the set of mechanical properties for each vertex, and the set of initial weights for each edge.

[0019] By employing the above technical solution, a set of mechanical attributes consisting of average force, peak force, and rate of change of force is bound to each vertex, enabling the graph structure to carry real-time stress state information of the nodes. Simultaneously, each edge is assigned static and dynamic weights calculated based on the vertex mechanical attributes, ensuring that edges not only represent connectivity but also reflect the load-bearing strength under static load and the force transmission stability under dynamic load. The vertex mechanical attribute set fully preserves the key information (average force, peak force, rate of change) extracted from the sliding window feature, while the edge weight set reflects load-bearing strength through static weights and force transmission stability through dynamic weights. Together, they constitute the multi-dimensional data input required for stress path identification. This structured data organization allows subsequent graph theory algorithms (such as connectivity verification and maximum connected component analysis) to be directly calculated based on the enhanced mechanical graph, avoiding the overhead of repeatedly querying the original sensor data and significantly improving the computational efficiency of the identification algorithm.

[0020] In a second aspect, embodiments of this application provide a control system for a distributed node scaffolding support system, including a memory, a processor, and a computer program stored in the memory; the processor executes the computer program to implement the steps of the method described in any possible implementation of the first aspect.

[0021] Thirdly, embodiments of this application provide a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the method described in any possible implementation of the first aspect.

[0022] Fourthly, embodiments of this application provide a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the method described in any possible implementation of the first aspect.

[0023] It is understood that the control system of the distributed node scaffolding support system provided in the second aspect, the storage medium provided in the third aspect, and the computer program product provided in the fourth aspect are all used to execute the method provided in this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By employing a sliding window algorithm locally on each intelligent node to extract three feature values ​​from the force data—average force, peak force, and rate of change of force—the massive amount of raw sampled data within each window is compressed into only three feature values. This processing allows each node to transmit and store only the feature values, rather than all the raw data, significantly reducing the bandwidth consumption of the communication medium in the distributed network. Simultaneously, it reduces the storage pressure on the microprocessors of each node, providing a feasible technical foundation for real-time collaborative control of large-scale scaffolding systems. 2. By abstracting the scaffolding structure into a graph structure with intelligent nodes as vertices and member connections as edges, the limitations of traditional methods that rely solely on stress magnitude thresholds for identification are overcome. By introducing the proportion of the maximum connected component, the topological connectivity of candidate main path nodes is verified, ensuring that the identified main load-bearing paths not only reach the stress threshold in terms of numerical values ​​but also form a continuous force transmission framework in space. This dual criterion effectively avoids the problem of isolated high-stress points being misidentified as main nodes, ensuring that the identification results truly reflect the force flow transmission path of the structure and providing a reliable decision-making basis for subsequent differentiated stiffness adjustments. 3. Mimicking the adaptive growth mechanism of mammalian skeletons, stiffness is enhanced at nodes along primary load-bearing paths (simulating the reinforcing growth of bones in stress concentration areas) and weakened at nodes along secondary load-bearing paths (simulating the absorption and thinning of bones in low-stress areas). Micrometer-level control of node stiffness is achieved by precisely adjusting the normal pressure at the node connection surfaces using piezoelectric stack actuators. This differentiated adjustment strategy actively reconstructs the force flow distribution of the structure, guiding loads to concentrate along the primary paths, enabling the structure to achieve optimal load-bearing efficiency with minimal material consumption. 4. Using the stress non-uniformity coefficient (coefficient of variation) as the quantitative convergence criterion, a closed-loop iterative architecture was established: "Learning period (sensing stress) → Analysis period (identifying the path) → Adaptive period (adjusting stiffness) → Convergence judgment." After each stiffness adjustment, stress data is re-collected to calculate the stress non-uniformity coefficient, and iteration stops when it is less than or equal to the preset convergence threshold. This mechanism enables the scaffolding structure to have continuous self-optimization capabilities, adapting to load changes and initial structural defects, and gradually approaching the optimal stress distribution state through multiple iterations. 5. Each intelligent node independently completes stress data acquisition, feature extraction, and stiffness adjustment locally, without relying on a central controller for centralized calculations, giving the system good robustness and scalability. Simultaneously, a distributed communication network enables the aggregation of feature data and collaborative adjustment among nodes. In the stress path identification phase, a global graph structure is constructed and connectivity verification is achieved. In the stiffness adjustment phase, influence ranking and connected component partitioning enable priority adjustment of key areas and intra-component collaborative synchronization. Attached Figure Description

[0025] Figure 1This is a flowchart illustrating the control method of the distributed node scaffolding support system in this application embodiment; Figure 2 This is a schematic diagram of the architecture of the smart node in the embodiments of this application; Figure 3 This is a schematic diagram illustrating the principle of the PID algorithm in the embodiments of this application; Figure 4 This is an exemplary hardware structure diagram of the control system of the distributed node scaffolding support system in the embodiments of this application. Detailed Implementation

[0026] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0027] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0028] The following is combined with Figure 1 The method of the embodiments of this application will be described below.

[0029] Please see Figure 1 This is a flowchart illustrating the control method of the distributed node scaffolding support system in this embodiment of the application. Figure 1 As shown, the control method includes the following steps: S101. After the scaffolding support system is erected and bears the load, the pressure sensors distributed on each intelligent node continuously collect the force data. The sliding window algorithm is used to extract features from the force data collected by each intelligent node to obtain feature data. The feature data includes the average force, peak force and rate of change of force for each window. S102. Collect the feature data of each intelligent node, convert the scaffolding support system into a graph structure with intelligent nodes as vertices and member connection relationships as edges, calculate the global mean and standard deviation of the average force of all intelligent nodes, set an initial force threshold, mark nodes whose average force reaches or exceeds the initial force threshold as candidate main path nodes, and verify the topological connectivity by calculating the proportion of the maximum connected component of the candidate main path nodes in the graph structure, mark the candidate main path nodes whose topological connectivity meets the preset requirements as main load-bearing path nodes, and mark the remaining nodes in the graph structure as secondary load-bearing path nodes; S103. For the first intelligent node marked as the primary load-bearing path node, control the piezoelectric stack actuator inside the first intelligent node to generate positive displacement, increase the positive pressure on the node connection surface to improve node stiffness. For the second intelligent node marked as the secondary load-bearing path node, control the piezoelectric stack actuator inside the second intelligent node to generate negative displacement or maintain zero displacement, reduce the positive pressure on the node connection surface to reduce node stiffness. S104. After each stiffness adjustment is completed, the stress data of each smart node is re-acquired to calculate the stress non-uniformity coefficient. When the stress non-uniformity coefficient is less than or equal to the preset convergence threshold, the iterative adjustment is stopped. The stress non-uniformity coefficient is used to characterize the coefficient of variation of the stress values ​​of all smart nodes.

[0030] After the scaffolding support system is erected and bears the initial load, the system enters the learning phase. Thin-film pressure sensors within each intelligent node continuously collect axial pressure data passing through that node at a preset sampling frequency (e.g., 100Hz). To avoid the communication burden caused by storing and transmitting massive amounts of raw data, each node uses a sliding window algorithm for feature extraction. A fixed window length (e.g., 60 seconds) can be set, with each window containing a fixed number of sampling points. The average force is obtained by summing the force values ​​of all sampling points within the window, reflecting the average load-bearing level of the node within the window period. The peak force is obtained by taking the maximum force value of all sampling points within the window, capturing instantaneous impacts or extreme load conditions. The rate of change of force is obtained by dividing the difference in force values ​​between the first and last sampling points of the window by the window duration, approximately reflecting the trend of force change (increase, decrease, or stability) within that period. After collecting the feature data from each intelligent node, the system enters the analysis phase, employing a stress path identification algorithm based on graph theory and threshold adaptation to distinguish between primary and secondary load-bearing paths in the structure. This process simulates the biological mechanism by which a skeleton defines its skeletal and auxiliary regions based on the magnitude of force and spatial continuity. The scaffolding system is abstracted as an undirected graph with intelligent nodes as vertices and member connections as edges. Each vertex is associated with the node's 3D coordinates and force characteristics to construct the graph structure. The global mean and standard deviation of the average force across all nodes are calculated. An initial force threshold is set as the sum of multiples of the mean and standard deviation. Nodes whose average force reaches or exceeds the threshold are marked as candidate primary path nodes. Topological connectivity verification: The induced subgraph formed by the candidate primary nodes is extracted, and a graph search algorithm is used to decompose the connected components. The proportion of nodes in the largest connected component to the total number of candidate nodes is calculated; this proportion reflects the degree to which the candidate nodes form a continuous force-transmitting skeleton in space. If connectivity does not meet the preset requirements, a bisection method is used to iteratively adjust the threshold within the force threshold range until connectivity meets the requirements, ultimately determining the primary and secondary load-bearing path node sets. Based on the identification results, the system enters an adaptive phase, performing differentiated stiffness adjustments on the primary and secondary path nodes. This process simulates the enhanced growth of bone in stress-concentrated areas and the absorption and thinning in low-stress areas. For smart nodes marked as primary load-bearing path nodes, the piezoelectric stack actuator is controlled to generate positive displacement. After being energized, the piezoelectric stack undergoes micron-level elongation due to the inverse piezoelectric effect, increasing the normal pressure between the node connection surfaces, thereby significantly increasing the maximum static friction of the connection surfaces, increasing the node stiffness, and enabling it to bear more loads. For smart nodes marked as secondary load-bearing path nodes, the piezoelectric stack actuator is controlled to generate negative displacement or maintain zero displacement, reducing or maintaining the normal pressure on the node connection surfaces, reducing the node stiffness, and actively unloading part of the load. A PID control algorithm is used to dynamically adjust the drive voltage based on the real-time force value feedback from the pressure sensor, so that the actual force on the node approaches the set target force value, achieving precise closed-loop adjustment of stiffness.After each stiffness adjustment, the system reassesses the overall stress distribution to determine if the optimization objective has been achieved. If not, it returns to step S101 to begin a new iteration until the structure converges to the optimal state. After each stiffness adjustment, the system enters a short relearning period to re-collect the stress data at each node. The stress non-uniformity coefficient (the ratio of standard deviation to mean) is calculated for all node stress values. This dimensionless coefficient objectively reflects the dispersion of the stress distribution throughout the structure. The stress non-uniformity coefficient is compared to a preset convergence threshold (ranging from 0.05 to 0.2). If the stress non-uniformity coefficient is less than or equal to the threshold, it indicates that the structural stress has become more uniform, and the system stops iterating and enters steady-state monitoring mode. If the stress non-uniformity coefficient is greater than the threshold, the system returns to step S101, using the adjusted stress state as the initial state for the new learning period, and continues the "learn-analysis-adjustment-evaluation" cycle.

[0031] This embodiment provides a biomimetic adaptive intelligent scaffolding system, which is assembled from multiple standard segments, and each connection node is an intelligent node. Figure 2 This is a schematic diagram of the architecture of the smart node in the embodiments of this application, such as... Figure 2As shown, the intelligent node includes a node housing 201, with connection interfaces 202 at both ends for connecting adjacent scaffolding members. The node housing 201 integrates a pressure sensor 203, a microprocessor 204, and a micro-electric expansion joint 205. The pressure sensor 203 is a thin-film pressure sensor embedded in the force-bearing surface of the node housing 201, used to detect the axial pressure passing through the node in real time. In this embodiment, the pressure sensor 203 operates at DC 5V, has a measurement range of 0-100kN, and an accuracy of 0.5%FS. The microprocessor 204 uses an ARM Cortex-M0 series low-power microcontroller with a main frequency of 48MHz, a built-in 12-bit ADC for acquiring pressure sensor signals, and multiple PWM output channels for controlling the micro-electric expansion joint. The microprocessor 204 also integrates a CAN controller for communication with adjacent nodes. The micro-electric expansion joint 205 uses a piezoelectric stack actuator, its core being a multi-layered piezoelectric ceramic sheet stack, which generates an inverse piezoelectric effect when energized, achieving micron-level expansion. The piezoelectric stack actuator used in this embodiment has dimensions of 5mm × 5mm × 10mm, a maximum stroke of 10μm, a maximum thrust of 500N, and a response time of less than 1ms. A miniature electric telescoping actuator 205 is installed at the connection surface of the node housing 201, with its extension direction perpendicular to the connection surface. The node stiffness is adjusted by changing the normal pressure on the connection surface. When the piezoelectric stack extends, the normal pressure on the connection surface increases, friction increases, and the node stiffness increases; conversely, the node stiffness decreases. The intelligent node also includes a communication unit 206, which uses a CAN bus with a communication rate of 1Mbps for data exchange between adjacent intelligent joint modules, forming a distributed intelligent network.

[0032] In some embodiments, the step of using a sliding window algorithm to extract features from the force data collected by each of the intelligent nodes to obtain feature data includes: The number of sampling points for each window is calculated using the following formula: ; in, N w Indicates the number of sampling points. L Indicates the window length. f s Indicates the sampling frequency; The average force for each window is calculated using the following formula: ; in, Indicates average force. F j Indicates the first in the window j The force values ​​at each sampling point; The peak force for each window is calculated using the following formula: ; in, F max Indicates peak force. F 1 The force value at the starting sampling point of the window. F Nw The force value at the end of the window sampling point. max Indicates the maximum value; The rate of change of force for each window is calculated using the following formula: ; in, r It represents the rate of change of force.

[0033] The product of the window length and the sampling frequency determines the data size of each window. Taking the embodiment of this application as an example, N w =60 × 100 = 6000 sampling points. This formula establishes a quantitative relationship between the time window and the amount of data, providing a unified window scale for subsequent feature extraction. The average force reflects the static bearing capacity of the node within the window period and is one of the core indicators for determining whether a node is on the main load-bearing path. By averaging 6000 sampling points, transient noise interference can be effectively eliminated, obtaining a stable force benchmark value. Peak force captures transient impacts or extreme loads occurring within the window period. During scaffolding construction, dynamic load impacts (such as material hoisting and personnel movement) may occur. These transient information may be smoothed out in the average force, but the peak force retains these key characteristics, providing an important basis for structural safety assessment. The rate of change of force reflects the trend of force change at the node within the window period. Positive values ​​indicate an increase in force, negative values ​​indicate a decrease in force, and the absolute value reflects the rate of change. This feature uses a simplified calculation method of dividing the difference between the beginning and end of the window by the window duration, avoiding complex linear fitting and significantly reducing computational complexity while ensuring sufficient accuracy. Pressure sensors continuously collect data on the corresponding skeleton's perception of the mechanical environment; sliding window feature extraction simulates the "sparse encoding" of sensory nerves for stimulus signals—transmitting only key features to the central nervous system, rather than all the original signals; the three feature values ​​of average force, peak force, and rate of change correspond to the skeleton's multi-dimensional memory of "steady-state load," "impact load," and "load trend," respectively. Through this biomimetic design, each intelligent node acts like a "sensory neuron," efficiently perceiving and encoding the mechanical environment at its location, providing high-quality input for the structure's self-optimization decisions.

[0034] In some embodiments, controlling the piezoelectric stack actuator inside the first smart node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Calculate the control voltage, and use the control voltage to control the piezoelectric stack actuator to generate a positive displacement, thereby increasing the positive pressure on the connecting surface; The control voltage is calculated using the following formula: ; in, e ( t )= F target - F cur , e ( t )express t The force error at any moment, F target Indicates the target force value. F cur express t Force value at all times u ( t )express t Control voltage at time, Δ t Indicates the control cycle. K p , K i , K d These represent the proportional, integral, and differential coefficients, respectively. e ( k ) indicates the first k Force error at each sampling time, k =0, 1, 2, ... t .

[0035] Control voltage u ( t The relationship between the piezoelectric stack actuator and the final node stiffness is established through the following chain: control voltage → inverse piezoelectric effect of the piezoelectric stack → micrometer-level displacement (elongation) → increased normal pressure on the connection surface → increased node stiffness → increased node force. The piezoelectric stack actuator consists of multiple layers of piezoelectric ceramic sheets stacked together. When a driving voltage is applied, an electric field is generated within the piezoelectric material, triggering the inverse piezoelectric effect: the material undergoes mechanical deformation in the direction of the electric field, resulting in elongation. The elongation is approximately proportional to the driving voltage. The piezoelectric stack is installed at the node connection surface, with its elongation direction perpendicular to the connection surface. When the piezoelectric stack elongates, it exerts a clamping effect on the connection surface, thereby increasing the normal pressure between the connection surfaces. Node stiffness is mainly manifested as the ability to resist relative rotation, and its magnitude depends primarily on the maximum static friction between the connection surfaces. When the normal pressure increases, the maximum static friction increases, enhancing the node's ability to resist relative slippage and rotation, i.e., increasing the node stiffness. F targetThe target force value is set as follows: for primary path nodes, the target force value is set to 1.2 times the current average force (20% increase); for secondary path nodes, the target force value is set to 0.5 times the current average force (50% decrease). F cur Indicates the current time t Real-time force values ​​acquired by a pressure sensor. Error. e ( t This reflects the difference between the force at the current node and the target force, and is the input signal of the PID controller: when e ( t When the stress is greater than 0, it indicates that the current stress is lower than the target, and the node stiffness needs to be increased to bear more load; when e ( t When the value is less than 0, it indicates that the current stress is higher than the target, and the node stiffness needs to be reduced to unload the load. K p , K i , K d These represent the proportional, integral, and differential coefficients, respectively, and their corresponding terms in the formula are the proportional term, integral term, and differential term. The proportional term represents the instantaneous value of the current error, used to quickly respond to the current deviation and determine the direction and magnitude of adjustment; the integral term represents the cumulative sum of historical errors, used to eliminate steady-state errors and ensure that the final force accurately reaches the target value; the differential term represents the rate of change of the error, used to predict the trend of error change and suppress overshoot and oscillation.

[0036] Elongation Δ of piezoelectric stack actuator L With driving voltage u ( t The relationship is approximately linear: Δ L =d 33 · n · u ( t ); where d 33 The piezoelectric strain constant is n The number of piezoelectric ceramic layers. Changes in control voltage are directly converted into micrometer-level displacement output. The elongation Δ of the piezoelectric stack actuator. L Apply normal pressure to the connection surface N Increase: N = N 0 + k s ·Δ L ;in N 0 As the initial normal force, k sThis represents the equivalent stiffness coefficient of the connection surface. Increased normal force leads to increased maximum static friction, thus increasing the node stiffness. With increased node stiffness, the node bears more load, and the node experiences increased stress. F cur Increase, error e ( t The pressure sensor collects new force values ​​in real time and feeds them back to the PID controller, forming a closed-loop regulation: sensing error → PID calculates voltage → piezoelectric stack elongates → stiffness increases → force increases → error decreases → voltage adjusts.

[0037] The piezoelectric stack actuator used in this application operates in a unipolar mode, with a driving voltage range of 0–150V (typical value, specific range depends on the model). When the driving voltage increases from 0V, the piezoelectric stack generates the inverse piezoelectric effect, elongating along the stacking direction and producing positive displacement; when the driving voltage decreases from the current value (including dropping to 0V), the piezoelectric stack contracts to a shorter length, producing negative displacement. That is, positive displacement corresponds to voltage increase, and negative displacement corresponds to voltage decrease, without the need to apply a negative voltage. Therefore, controlling the piezoelectric stack actuator to produce positive displacement means increasing the driving voltage, and controlling the piezoelectric stack actuator to produce negative displacement means decreasing the driving voltage (including complete de-energization). By adjusting the magnitude of the driving voltage, the displacement can be continuously controlled, thereby changing the normal pressure on the connection surface and achieving bidirectional adjustment of the node stiffness. In one embodiment of this application, the rated driving voltage of the piezoelectric stack actuator is 150V, and the initial state is 0V (fully contracted). When increased stiffness is required, the voltage is increased to 120V~150V to obtain positive displacement; when decreased stiffness is required, the voltage is reduced to 0V~50V to cause the piezoelectric stack to shrink back to its initial or near-initial length.

[0038] Figure 3 This is a schematic diagram illustrating the principle of the PID algorithm in the embodiments of this application, as shown below. Figure 3 As shown, the data input process involves collecting the average force, coordinates, and connection relationships of all nodes; statistical calculations are performed to calculate the global mean *a* and standard deviation *b*; an initial threshold is set, with threshold *c* = *a* + *kb*, where *k* = 1.5; and preliminary partitioning is performed, dividing P based on the threshold. candidate and S candidate Construct a subgraph, with P candidate Construct subgraph G for vertices P Connectivity calculation to find the largest connected component C. max Calculate the proportion η; determine connectivity: η≥0.9? If yes, accept the classification and determine the final primary node set P and secondary node set S; assign roles to each node, labeling it MAJOR (primary) or MINOR (minor); broadcast the results to notify neighboring nodes; output the results, outputting the node roles for use during the adaptation period; if no, adjust the threshold, updating the threshold using a binary search method.

[0039] In some embodiments, controlling the piezoelectric stack actuator inside the first smart node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Based on the topological connection relationship of the scaffolding support system, the entire topological structure is divided into multiple connected components, and the intelligent nodes in each connected component are physically connected to each other. For each connected component, an influence factor is calculated based on the force weight of each smart node and the number of connecting members in the structural topology. The force weight is the proportion of the node's force value to the total force of the structure, and the number of connecting members represents the importance of the node in the structure. The connected components are prioritized according to the order of their influence factors from largest to smallest. Based on the priority order, stiffness adjustment is performed on each connected component in turn, starting from the connected component with the largest influence factor. For each of the intelligent nodes within the same connected component, a local synchronization signal is generated through a distributed communication network to simultaneously perform stiffness adjustment within the same time window. During the adjustment process, the stiffness adjustment amount of each of the intelligent nodes is multiplied by a dynamic damping factor, which gradually increases over time. During stiffness adjustment, the rate of force change of each intelligent node and the amplitude of force fluctuation between adjacent intelligent nodes are monitored in real time. If the rate of change of force exceeds the preset safety rate threshold, or the amplitude of force fluctuation at adjacent nodes exceeds the preset fluctuation threshold, the current adjustment is paused, the control voltage is restored to the state before adjustment, and the adjustment is re-executed with a reduced adjustment step size after the stress wave dissipates.

[0040] Connected components are a fundamental concept in graph theory, referring to the largest subgraph in an undirected graph where every two vertices are connected by a path. In scaffolding structures, connected components correspond to physically interconnected groups of nodes, with no direct link connections between different components. Based on the graph structure constructed in step S102, depth-first search or breadth-first search algorithms are used to identify all connected components. Nodes within each connected component are directly or indirectly connected by links, forming a continuous force transmission region. The influence factor comprehensively reflects the mechanical and topological importance of nodes in the structure: the proportion of the node's force value to the total structural force reflects the node's contribution to load transfer; nodes with a larger force weight bear more loads and have a greater impact on the overall stress distribution. The degree of a node in the structural topology (i.e., the number of links connected to that node) reflects the node's topological importance in the structure. Nodes with more connecting links are often the "hubs" of the structure, and their stiffness changes have a wider impact on surrounding nodes. Comprehensive calculation method: Influence factor = Force weight × (1 + α × Number of connecting members / Maximum number of connecting members); where α is the weighting coefficient (range 0.2~0.5), used to balance the contributions of mechanical importance and topological importance. Connected components are arranged in descending order of influence factor, with the component with the highest influence at the top. After the current component completes all stiffness adjustments and its stress state stabilizes, the adjustment of the next priority component is initiated. The connected component with the highest influence bears the most important load transfer function in the structure; prioritizing its adjustment can improve the overall stress distribution most quickly. Sequential execution between components effectively avoids global coupling oscillations caused by simultaneous adjustments of multiple components, ensuring the stability of the control process. Nodes within the same connected component exchange status information through distributed communication networks such as CAN bus. Through local synchronization signals, stiffness adjustments are initiated simultaneously within the same time window, ensuring the consistency of the adjustment sequence of nodes within the component and avoiding mutual constraints caused by differences in adjustment sequence. Dynamic damping factor design: The damping factor is small at the initial moment (e.g., 0.6), gradually increasing over time (e.g., to 1.0). Adjustment Calculation: Actual adjustment = Target adjustment × γ(t), where γ(t) is the dynamic damping factor that increases with time. Time Constant Setting: The growth rate of the damping factor is determined by the time constant, giving the adjustment process a "slow at first, then rapid" characteristic. Force Change Rate Monitoring: Calculates the change in force at the nodes per unit time, reflecting the severity of the adjustment process. Adjacent Node Force Fluctuation Monitoring: Monitors the change in force difference between adjacent nodes, reflecting the degree of influence of the adjustment on surrounding nodes. Monitoring Frequency: Synchronized with the control cycle, acquiring monitoring data in real time.If the rate of change of force exceeds a preset safety threshold (e.g., 100 N / s) or the amplitude of force fluctuation at adjacent nodes exceeds a preset fluctuation threshold (e.g., ±15%), the current adjustment action should be stopped immediately. The drive voltage of the piezoelectric stack actuator should be restored to its pre-adjustment state to restore the node stiffness to its original state. A preset time (determined based on the stress wave propagation speed) should be waited to ensure sufficient attenuation of the mechanical disturbance before readjusting with a reduced adjustment step size (e.g., 70%~80% of the original step size). In structural mechanics, when the stiffness of a node changes, stress waves are generated and propagate along the members to the surrounding area. The propagation speed of the stress wave depends on the elastic modulus and density of the material. This embodiment requires waiting for the stress wave to dissipate after rollback to ensure sufficient attenuation of the mechanical disturbance before readjustment.

[0041] In some embodiments, the re-collection of the stress data of each of the smart nodes to calculate the stress non-uniformity coefficient includes: The stress non-uniformity coefficient is calculated using the following formula: ; ; in, This represents the stress non-uniformity coefficient, and m represents the total number of smart nodes. Let μ be the average force value of the i-th node, and μ be the global mean of the average force of all smart nodes.

[0042] The global mean reflects the average stress level of the entire structure under the current load state and is the benchmark reference value for the stress values ​​of all nodes. Regardless of the size of the structure or the absolute value of the load, the global mean provides a unified scale for subsequent calculations. The term within the square root is the variance of the stress values ​​of all nodes, reflecting the degree of dispersion of the stress of each node relative to the global mean; the standard deviation is obtained after taking the square root, which represents the average fluctuation range of the stress value; the coefficient of variation, i.e., the stress non-uniformity coefficient, is obtained by dividing the standard deviation by the global mean. The stress non-uniformity coefficient is the coefficient of variation in statistics and has the following characteristics: (1) Dimensionless: After dividing by the mean, the influence of dimensions is eliminated, and it is not limited by the size of the structure or the absolute value of the load; (2) Standardized comparison: The uniformity of stress distribution under different structures and different load conditions can be compared using a unified index; (3) The smaller the value, the closer the stress of each node is to the average value and the more uniform the stress distribution; the larger the value, the more uneven the stress distribution, and there is obvious stress concentration or local unloading phenomenon. Preferably, the range of the preset convergence threshold can be [0.05, 2].

[0043] In some embodiments, verifying topological connectivity by calculating the percentage of the largest connected component in the graph structure of the candidate major path nodes includes: Extract the induced subgraph from the graph structure, which is composed of the candidate main path nodes and the connecting edges between the candidate main path nodes. Use a graph search algorithm to decompose the induced subgraph into connected components, identify all connected components contained in the induced subgraph, and record the node information contained in each connected component. From all connected components, the connected component with the most nodes is selected as the maximum connected component. The relative proportion of the number of nodes contained in the maximum connected component to the total number of candidate main path nodes is calculated, and the relative proportion is compared with the preset connectivity requirements. If the relative proportions meet the preset connectivity requirements, then the candidate main path nodes are determined to satisfy the topological connectivity conditions, and the current classification result is accepted. If the relative proportion does not meet the preset connectivity requirements, the candidate main path node is determined to not meet the connectivity conditions. The initial force threshold is iteratively adjusted within the force threshold range using a binary search method until the connectivity index meets the preset requirements or reaches the maximum number of iterations.

[0044] Based on the graph structure G=(V,E) constructed in step S102, where V represents all smart nodes and E represents the link connections between nodes, let P be the set of candidate main path nodes. candidate Extract from G the P candidate The induced subgraph G formed by the edges connecting them. P Induced subgraph G P This only includes candidate principal nodes and their existing connections. By extracting the induced subgraph, the analysis scope is narrowed from the entire graph to the candidate principal nodes and their relationships, providing a precise computational object for subsequent connectivity analysis and avoiding redundant calculations across the entire graph. Depth-first search or breadth-first search algorithms can be used to traverse the induced subgraph. In an undirected graph, a connected component is the largest subgraph where any two vertices are connected by a path. The number and composition of nodes in each connected component are recorded. Connected component decomposition divides the candidate principal nodes into several independent connected groups. Each connected component represents a group of spatially connected candidate principal nodes, potentially forming local force transmission paths. From all connected components, the connected component with the largest number of nodes is selected as the largest connected component. The largest connected component represents the most concentrated group of candidate principal nodes and has the greatest potential to form the main force transmission path of a through structure. Let the total number of candidate principal path nodes be |P|. candidate |, the maximum number of nodes in the connected component is |C max |, connectivity index η=|C max ∣ / ∣P candidate| Connectivity index reflects the degree to which candidate main nodes form a continuous force transmission skeleton in space: η close to 1: the vast majority of candidate main nodes are concentrated in the largest connected component, forming a highly connected main force transmission skeleton; η significantly less than 1: candidate main nodes are scattered in multiple isolated small components, with broken force flow paths. Preset connectivity requirement: In this embodiment, η is set to ≥ 0.9, that is, the proportion of the largest connected component reaches more than 90%. If η ≥ 0.9, the candidate main nodes are determined to meet the topological connectivity condition, and the current classification result based on the force threshold is accepted; if η < 0.9, the candidate main nodes are determined to be too scattered, with isolated high-stress nodes or broken force flow paths, and the identification parameters need to be adjusted. When η = 1, all candidate main nodes are completely connected, forming a single force transmission skeleton; when η < 1, there are isolated nodes or branches. Theoretical analysis shows that when η is below 0.85, there are multiple small components outside the maximum connected component. These small components often correspond to nodes with high stress but isolated locations (such as cantilever ends and local concentrated load points). If these are mistakenly identified as main nodes and their stiffness is increased, it will lead to incorrect stiffness distribution. Through a large number of finite element simulations of scaffolding structures, the proportion of nodes that truly bear the main force transmission function to the total number of high-stress nodes under different load conditions was statistically analyzed. It was found that the lower limit of this proportion is about 0.88~0.92. Therefore, η≥0.9 is taken as the safety boundary. When η<0.9, it indicates that the candidate main nodes divided by the current stress threshold are not spatially concentrated enough. The main reasons may be: the threshold is set too low, causing too many nodes with average stress to be included in the candidate set, including isolated nodes; the threshold is set too high, causing nodes that truly constitute the force transmission path to be excluded, leaving only a few isolated high-stress points. A bisection method is used to iteratively adjust the initial stress threshold within the stress threshold range until the connectivity index reaches the preset requirement or the maximum number of iterations is reached. Lower limit T of the threshold low =μ (global mean), upper threshold (The maximum average force across all nodes). Bisection iterative process: Calculate the current threshold T = (T... low +T high ) / 2; Re-divide the candidate principal node set P according to the new threshold T. candidate Recalculate the connectivity index η; adjust the threshold range based on η. If η < 0.9, the current threshold is still not ideal and needs further adjustment. If too many candidate nodes lead to dispersion, increase the lower limit T of the threshold. low =T; If there are too few candidate nodes, making it impossible to form a connection, then lower the upper threshold T. high =T. Repeat the above steps until η≥0.9 or the maximum number of iterations is reached. Iteration termination condition: Successful termination: η≥0.9, the optimal threshold T is obtained. optBased on this, the final set of primary path nodes is determined; forced termination: upon reaching the maximum number of iterations (e.g., 10), the threshold with the best connectivity in the current iteration is used as the final result. After iterative adjustment, the final set of primary and secondary load-bearing path nodes is determined based on the optimal threshold. Primary load-bearing path node set: average force value ≥T opt And the nodes that belong to the largest connected component; the set of secondary load-bearing path nodes: the remaining nodes in the graph structure.

[0045] In some embodiments, converting the scaffolding support system into a graph structure with smart nodes as vertices and member connection relationships as edges includes: Based on the design drawings or actual assembly information of the scaffolding support system, each smart node is mapped to a vertex in the graph structure, and the connection relationship between adjacent smart nodes is mapped to an edge in the graph structure. An initial undirected graph reflecting the physical connection relationship of the structure is constructed, and three-dimensional coordinate information in space is associated with each vertex. Based on the force characteristic data collected by each intelligent node, a set of mechanical attributes is bound to each vertex in the initial undirected graph. The set of mechanical attributes includes the average force value, peak force value, and force change rate corresponding to the intelligent node. The static weight is determined by the sum of the average force values ​​of the two vertices connected to each edge in the initial undirected graph, and the dynamic weight is determined by the difference in the rate of change of the force of the two vertices connected to each edge. An initial weight is assigned to each edge in the initial undirected graph based on the static weight and the dynamic weight. The graph structure is constructed based on the set of vertices, the set of edges, the set of mechanical properties for each vertex, and the set of initial weights for each edge.

[0046] Vertex Mapping: Each intelligent node in the scaffolding support system is mapped to a vertex in the graph structure. Each vertex represents an intelligent node with sensing, processing, and execution capabilities. The number of vertices equals the total number of intelligent nodes. Edge Mapping: The connection relationship between adjacent intelligent nodes is mapped to an edge in the graph structure. Edges represent physical connections between nodes, including connections between uprights and horizontals, and connections between diagonal braces. Since the connections between members in the scaffolding structure are usually bidirectional, an undirected graph model is used. 3D Coordinate Association: Each vertex is associated with its 3D coordinate information in space. The coordinate information can be obtained from design drawings or through positioning measurements during actual assembly. The association of 3D coordinates provides the geometric basis for subsequent spatial analysis and connectivity verification. Mechanical property data comes from the feature data extracted by the sliding window algorithm in step S101. After each intelligent node completes feature extraction locally, its feature values ​​are uploaded or used for graph structure construction. The mechanical property set includes average force value, peak force value, and force change rate. Each vertex is associated with a mechanical property vector, which includes average force value, peak force value, and force change rate. After attribute binding, each vertex carries the real-time mechanical state information of that node. The static weight is determined based on the sum of the average forces on the two vertices connected by each edge, calculated using the following formula: ;in and They are the edges e ij The two vertices connected v i and v j The average stress value. Static weight reflects the load-bearing strength of the connection under static load. The greater the sum of the average forces at the two endpoints, the higher the static weight, and the more likely the connection is to be part of the main force transmission path. Dynamic weight is determined by the absolute value of the difference in the rate of change of force between the two vertices connected by each edge. The calculation formula is: w dynamic ( e ij = | r i - r j |; among which r i and r j They are the edges e ij The two vertices connected v i and v jThe dynamic weight reflects the stress transfer stability of the connection under dynamic loads. The smaller the difference in the rate of change of stress at the two endpoints, the better the synchronicity of stress changes at both ends and the smoother the stress transfer. The lower the dynamic weight, the more stable the connection is under dynamic loads and the more suitable it is as a stable force transmission path. The initial weight of the edge is determined by combining the static and dynamic weights. Edges with higher static weights and lower dynamic weights should be assigned higher initial weights, indicating that the connection is both a high-load-bearing area and has good force transmission stability. The initial weight can be determined by weighted summation or other comprehensive evaluation methods. The vertex set includes all smart nodes, the edge set includes all rod connection relationships, the vertex mechanical attribute set includes the average stress value, peak stress value, and rate of change of stress bound to each vertex, and the edge weight set includes the static and dynamic weights of each edge. The graph structure is constructed based on the vector representation of the vertex set, edge set, mechanical attribute set of each vertex, and initial weight set of each edge. In subsequent iterations, after each intelligent node completes stiffness adjustment and enters a new learning period, the above steps are repeated based on the newly collected force characteristic data to dynamically update the vertex mechanical properties and edge weights in the mechanical enhancement graph.

[0047] The control method of the scaffolding support system of distributed nodes in the embodiments of this application has been described above. The control system of the scaffolding support system of distributed nodes in the embodiments of this application will be described in detail below in conjunction with the control method of the scaffolding support system of distributed nodes described above.

[0048] Please see Figure 4 This is an exemplary hardware structure diagram of the control system of the distributed node scaffolding support system in this application embodiment.

[0049] In some embodiments, the control system 400 of the distributed node scaffolding support system includes a computer device, which may be a terminal device. The computer device includes a processor 401, a memory 402, a sensor module 403, a communication module 404, an input device 405, and an output device 406 connected via a system bus. The processor 401 provides computing and control capabilities. The memory 402 includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data.

[0050] Those skilled in the art will understand that Figure 4The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0051] In some embodiments of this application, a computer-readable storage medium is provided, including instructions that, when executed on the control system 400 of the scaffolding support system of the distributed node, cause the control system 400 of the scaffolding support system of the distributed node to execute the control method of the scaffolding support system of the distributed node in the embodiments of this application.

[0052] In some embodiments of this application, a computer program product is also provided. When the computer program product is run on the control system 400 of the scaffolding support system of the distributed node, the control system 400 of the scaffolding support system of the distributed node executes the control method of the scaffolding support system of the distributed node in the embodiments of this application.

[0053] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0054] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0055] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A control method for a distributed node scaffolding support system, characterized in that, The scaffolding support system includes multiple intelligent nodes, and the methods include: After the scaffolding support system is erected and bears the load, pressure sensors distributed on each intelligent node continuously collect force data. The sliding window algorithm is used to extract features from the force data collected by each intelligent node to obtain feature data. The feature data includes the average force, peak force and rate of change of force for each window. The feature data of each intelligent node are collected, and the scaffolding support system is converted into a graph structure with intelligent nodes as vertices and member connection relationships as edges. The global mean and standard deviation of the average force of all intelligent nodes are calculated, an initial stress threshold is set, and nodes whose average force reaches or exceeds the initial stress threshold are marked as candidate main path nodes. The topological connectivity is verified by calculating the proportion of the maximum connected component of the candidate main path nodes in the graph structure. The candidate main path nodes whose topological connectivity meets the preset requirements are marked as main load-bearing path nodes, and the remaining nodes in the graph structure are marked as secondary load-bearing path nodes. For the first smart node marked as the primary load-bearing path node, the piezoelectric stack actuator inside the first smart node is controlled to generate positive displacement, increasing the positive pressure on the node connection surface to improve node stiffness. For the second smart node marked as the secondary load-bearing path node, the piezoelectric stack actuator inside the second smart node is controlled to generate negative displacement or maintain zero displacement, reducing the positive pressure on the node connection surface to reduce node stiffness. After each stiffness adjustment is completed, the stress data of each smart node is re-acquired to calculate the stress non-uniformity coefficient. When the stress non-uniformity coefficient is less than or equal to the preset convergence threshold, the iterative adjustment is stopped. The stress non-uniformity coefficient is used to characterize the coefficient of variation of the stress values ​​of all smart nodes.

2. The method according to claim 1, characterized in that, The feature data obtained by using the sliding window algorithm to extract features from the force data collected by each of the intelligent nodes includes: The number of sampling points for each window is calculated using the following formula: ； in, N w Indicates the number of sampling points. L Indicates the window length. f s Indicates the sampling frequency; The average force for each window is calculated using the following formula: ； in, Indicates average force. F j Indicates the first in the window j The force values ​​at each sampling point; The peak force for each window is calculated using the following formula: ； in, F max Indicates peak force. F 1 The force value at the starting sampling point of the window. F Nw The force value at the end of the window sampling point. max Indicates the maximum value; The rate of change of force for each window is calculated using the following formula: ； in, r It represents the rate of change of force.

3. The method according to claim 1, characterized in that, The method of controlling the piezoelectric stack actuator inside the first intelligent node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Calculate the control voltage, and use the control voltage to control the piezoelectric stack actuator to generate a positive displacement, thereby increasing the positive pressure on the connecting surface; The control voltage is calculated using the following formula: ； in, e ( t )= F target - F cur , e ( t )express t The force error at any moment, F target Indicates the target force value. F cur express t Force value at all times u ( t )express t Control voltage at any time, Δ t Indicates the control period. K p , K i , K d These represent the proportional, integral, and differential coefficients, respectively. e ( k ) indicates the first k Force error at each sampling time, k =0, 1, 2, ... t .

4. The method according to claim 3, characterized in that, The method of controlling the piezoelectric stack actuator inside the first intelligent node to generate positive displacement, increasing the normal pressure on the node connection surface to improve node stiffness, specifically includes: Based on the topological connection relationship of the scaffolding support system, the entire topological structure is divided into multiple connected components, and the intelligent nodes in each connected component are physically connected to each other. For each connected component, an influence factor is calculated based on the force weight of each smart node and the number of connecting members in the structural topology. The force weight is the proportion of the node's force value to the total force of the structure, and the number of connecting members represents the importance of the node in the structure. The connected components are prioritized according to the order of their influence factors from largest to smallest. Based on the priority order, stiffness adjustment is performed on each connected component in turn, starting from the connected component with the largest influence factor. For each of the intelligent nodes within the same connected component, a local synchronization signal is generated through a distributed communication network to simultaneously perform stiffness adjustment within the same time window. During the adjustment process, the stiffness adjustment amount of each of the intelligent nodes is multiplied by a dynamic damping factor, which gradually increases over time. During stiffness adjustment, the rate of force change of each intelligent node and the amplitude of force fluctuation between adjacent intelligent nodes are monitored in real time. If the rate of change of force exceeds the preset safety rate threshold, or the amplitude of force fluctuation at adjacent nodes exceeds the preset fluctuation threshold, the current adjustment is paused, the control voltage is restored to the state before adjustment, and the adjustment is re-executed with a reduced adjustment step size after the stress wave dissipates.

5. The method according to claim 1, characterized in that, The process of re-collecting the stress data of each of the smart nodes to calculate the stress non-uniformity coefficient includes: The stress non-uniformity coefficient is calculated using the following formula: ； ； in, This represents the stress non-uniformity coefficient, and m represents the total number of smart nodes. Let μ be the average force value of the i-th node, and μ be the global mean of the average force of all smart nodes.

6. The method according to claim 1, characterized in that, The step of verifying topological connectivity by calculating the percentage of the largest connected component in the graph structure of the candidate major path nodes includes: Extract the induced subgraph from the graph structure, which is composed of the candidate main path nodes and the connecting edges between the candidate main path nodes. Use a graph search algorithm to decompose the induced subgraph into connected components, identify all connected components contained in the induced subgraph, and record the node information contained in each connected component. From all connected components, the connected component with the most nodes is selected as the maximum connected component. The relative proportion of the number of nodes contained in the maximum connected component to the total number of candidate main path nodes is calculated, and the relative proportion is compared with the preset connectivity requirements. If the relative proportions meet the preset connectivity requirements, then the candidate main path nodes are determined to satisfy the topological connectivity conditions, and the current classification result is accepted. If the relative proportion does not meet the preset connectivity requirements, the candidate main path node is determined to not meet the connectivity conditions. The initial force threshold is iteratively adjusted within the force threshold range using a binary search method until the connectivity index meets the preset requirements or reaches the maximum number of iterations.

7. The method according to claim 1, characterized in that, The process of converting the scaffolding support system into a graph structure with smart nodes as vertices and member connection relationships as edges includes: Based on the design drawings or actual assembly information of the scaffolding support system, each smart node is mapped to a vertex in the graph structure, and the connection relationship between adjacent smart nodes is mapped to an edge in the graph structure. An initial undirected graph reflecting the physical connection relationship of the structure is constructed, and three-dimensional coordinate information in space is associated with each vertex. Based on the force characteristic data collected by each intelligent node, a set of mechanical attributes is bound to each vertex in the initial undirected graph. The set of mechanical attributes includes the average force value, peak force value, and force change rate corresponding to the intelligent node. The static weight is determined by the sum of the average force values ​​of the two vertices connected to each edge in the initial undirected graph, and the dynamic weight is determined by the difference in the rate of change of the force of the two vertices connected to each edge. An initial weight is assigned to each edge in the initial undirected graph based on the static weight and the dynamic weight. The graph structure is constructed based on the set of vertices, the set of edges, the set of mechanical properties for each vertex, and the set of initial weights for each edge.

8. A control system for a distributed node scaffolding support system, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-7.

Images