A method for monitoring the safety of a mold frame of an ultra-high and large-span warehouse cross beam structure

Abstract

The application discloses a kind of superhigh large-span cold store well beam structure's mould frame safety monitoring method, comprising the following steps: establishing digital force transmission topology diagram;Sensor is laid out in topological diagram key node, synchronous acquisition axial force time series data;The correlation coefficient of adjacent node axial force time series is calculated, and the correlation matrix indicating the close degree of force transmission is constructed;Determine neighborhood node group based on correlation matrix, calculate the deviation degree of each node axial force relative to neighborhood weighted average value;Abnormal node is identified in combination with regional differentiation threshold value;Based on the linkage intensity of correlation matrix, the abnormal node is taken as the starting point, and the path is terminated to the top support layer well beam rib projection boundary along the load transfer direction of topological edge inverse load in trace, and the load disturbance input point is recorded as the path termination vertical rod, and the abnormal load variation area is defined;Report the area.The present application realizes accurate positioning from single-point anomaly to regional source by topological inverse tracing, which provides basis for construction safety control.

Description

Technical Field

[0001] This invention relates to the field of building construction technology, specifically to a method for safety monitoring of the formwork of a super-high, large-span cold storage grid beam structure. Background Technology

[0002] With the rapid development of the cold chain logistics industry, ultra-high and large-span cold storage buildings are becoming increasingly common. These cold storage facilities typically have large storage spaces and high floor heights, and their roof structures often adopt a large-span grid beam design to balance load-bearing requirements with the building's spatial layout.

[0003] During concrete pouring, the formwork system bears multiple loads, including the weight of the freshly poured concrete, the weight of the formwork itself, construction live loads, and pumping impact loads. These loads are transferred to the uprights via the top supports, and then downwards layer by layer through a force transmission network composed of horizontal bars and scissor braces. In this process, the area below the ribs of the grid beams becomes the main area of ​​concentrated load due to the large cross-section and heavy weight of the beams. The axial force on the uprights is much higher than in the non-ribbed areas, resulting in significant differences in load distribution. Under these complex stress conditions, any abnormal stress in any upright or member of the formwork system can spread to the surrounding area through the connection nodes. If this is not detected and addressed in time, it may trigger a chain reaction, ultimately leading to local instability or even the collapse of the entire structure.

[0004] Currently, common practices for safety monitoring of tall formwork support systems include installing sensors such as strain gauges, axial force gauges, and displacement gauges at key locations to collect parameters such as the axial force of the uprights, formwork settlement, and horizontal displacement of the frame in real time, and triggering an alarm when a parameter exceeds a preset threshold.

[0005] However, when the aforementioned existing monitoring technologies are applied to the formwork system of ultra-high, large-span cold storage grid beam structures, the monitoring indicators are independent of each other, lacking a comprehensive assessment of the overall stress state of the formwork system. Existing methods set alarm thresholds for single physical quantities, failing to correlate and analyze monitoring data from adjacent nodes, making it difficult to reflect the load transmission path and diffusion range within the formwork system. When an abnormal stress occurs at a node, existing methods can only report the node's over-limit state, unable to determine whether the abnormality stems from insufficient load-bearing capacity of the node itself or from load transfer from surrounding areas.

[0006] Meanwhile, as a spatial structure composed of numerous members connected by nodes, the formwork system exhibits significant transmissibility and diffusion of stress states. Isolated point monitoring is insufficient to cover the entire process of load anomalies propagating along the topological path. When the monitoring system detects an anomaly, existing technologies can only provide information on the location of the exceeding limit, failing to trace the source of the load disturbance or define the affected structural area. This makes it difficult for on-site management personnel to quickly pinpoint the root cause and take targeted measures. Therefore, it is necessary to research a formwork safety monitoring method for ultra-high, large-span cold storage grid beam structures. Summary of the Invention

[0007] Therefore, the purpose of this invention is to provide a method for safety monitoring of formwork in ultra-high, large-span cold storage grid beam structures, which effectively solves the problem that existing formwork monitoring methods can only perform single-point alarms and cannot trace the source of abnormal load disturbances, making it difficult to locate the hidden danger area and its cause on site.

[0008] To achieve the above objectives, the technical solution adopted by this invention is: a method for safety monitoring of the formwork of an ultra-high, large-span cold storage grid beam structure, comprising the following steps:

[0009] S1. Based on the geometric topological relationship and component connection properties of the formwork system, establish a digital force transmission topology diagram;

[0010] S2, Deploy a sensor network at the key nodes of the topology to synchronously collect axial force time-series data of each node;

[0011] S3, based on the axial force time series data of neighboring nodes, calculate the load transfer correlation coefficient between nodes and construct an association matrix representing the degree of force transfer between nodes;

[0012] S4. Based on the correlation matrix, determine the neighboring node groups with a preset topological distance centered on each node, and calculate the degree of axial force deviation of each node.

[0013] S5, predefine safety thresholds for different regions. When the axial force deviation of a node exceeds the preset safety threshold of the corresponding region, the node is marked as an abnormal node.

[0014] S6. Based on the linkage strength of the correlation matrix, starting from the abnormal node, trace back along the topological edge against the load transfer direction to the projection boundary of the grid beam rib of the top support layer, and record the terminal column of the path on the projection boundary as the load disturbance input point; define the grid beam rib segment and adjacent plate area to which the load disturbance input point belongs as the abnormal load change area.

[0015] S7, report areas with abnormal load changes.

[0016] Furthermore, a digital force transmission topology map is constructed, including:

[0017] The uprights with axial force bearing capacity in the formwork system are taken as topological vertices, and the horizontal bars and scissor braces are taken as topological edges connecting the vertices. The spatial coordinates of the vertices and the cross-sectional mechanical parameters of the topological edges are entered according to the design drawings, and the stiffness weights of each topological edge are obtained according to the cross-sectional mechanical parameters.

[0018] Based on the node construction attributes of fastener-type or disc-type nodes, determine the connection constraints between vertices.

[0019] By integrating the spatial coordinates, stiffness weights, and connection constraints, a digital force transmission topology diagram containing geometric topology information and mechanical transmission weights is generated.

[0020] Furthermore, construct the association matrix, including:

[0021] Extract the axial force time series data of each neighboring node within the same monitoring window, calculate the similarity of the axial force fluctuation curves of any two nodes, and use the similarity as the linkage strength between nodes;

[0022] The linkage strength between each node is normalized, and the association matrix is ​​generated according to the connection relationship of the topological vertices.

[0023] Furthermore, based on a preset time step, the linkage strength between nodes is calculated and the correlation matrix is ​​updated.

[0024] Furthermore, taking the target node as the center, all nodes within the first-order neighborhood of the topological distance in the correlation matrix are extracted to form a neighborhood node group;

[0025] Based on the linkage strength values ​​in the correlation matrix, the weighted average value of the axial force monitoring values ​​of the neighboring node group is calculated to obtain the neighborhood axial force benchmark value; the absolute value of the difference between the target node's axial force monitoring value and the neighborhood axial force benchmark value is used as the degree of axial force deviation.

[0026] Furthermore, a first safety threshold is set for the nodes located below the projection of the ribs of the grid beam;

[0027] For nodes located in the non-ribbed area of ​​the grid beam, a second safety threshold is set, wherein the second safety threshold is less than the first safety threshold;

[0028] When the deviation of the axial force exceeds the safety threshold of the corresponding area, the marking process of the abnormal node is triggered.

[0029] Furthermore, when there are multiple marked abnormal nodes, the observation time window is set with the first time the first abnormal node exceeds the limit as the starting point; the node with the largest axial force deviation within the observation time window is taken as the starting point.

[0030] Furthermore, weak connection edges are filtered out based on the linkage strength of the correlation matrix. Starting from the abnormal node, a recursive search is performed along the strong connection edge towards the topological boundary, and the sequence of nodes passed along the way is recorded as a candidate transmission path.

[0031] Calculate the cumulative value of linkage intensity in each candidate transmission path, and determine the candidate transmission path with the largest cumulative value as the main transmission path.

[0032] Furthermore, the vertical deformation time series data of each node are collected synchronously; based on the linkage intensity value in the correlation matrix, the weighted average value of the vertical deformation time series data of the neighboring node group is calculated to obtain the neighborhood deformation reference value; the absolute value of the difference between the vertical deformation time series data of the target node and the neighborhood deformation reference value is calculated and defined as the degree of deformation deviation; the ratio of the degree of axial force deviation to the degree of deformation deviation is calculated as the deviation coefficient of the node.

[0033] If the deviation coefficient is within the first preset range, the node is determined to be normal, and S6 is performed.

[0034] If the deviation coefficient is greater than the first preset interval, the node is determined to be degraded, and a structural anomaly alarm containing location information is reported.

[0035] Furthermore, a sample library of load transfer paths for formwork under different failure modes is constructed based on finite element simulation software, and a mapping relationship between load transfer path feature vectors and abnormal regions is established. The main transfer paths determined in real time are converted into feature vectors, and similarity matching is performed in the sample library to verify and correct the definition results of the abnormal load variation regions.

[0036] The beneficial effects of the above technical solution are as follows: For the safety monitoring scenario of the formwork support system in the concrete pouring stage of a super-high, large-span cold storage grid roof structure, the load levels in the beam rib area and slab strip area differ significantly, and the force on the uprights exhibits obvious spatial transfer and diffusion. Traditional monitoring methods often rely on fixed threshold alarms for single points and single physical quantities, which are insufficient to reflect the overall stress state of the formwork and the load transfer path in the support system, and cannot effectively locate the source of load disturbance and the affected area. Therefore, this invention proposes a monitoring method based on a digital force transmission topology diagram, node correlation matrix, joint analysis of axial force and vertical deformation, and finite element sample library matching. In principle, this method elevates discrete point monitoring results to the diagnosis and verification of abnormal load areas and sources.

[0037] This invention uses vertical poles with axial load-bearing capacity as topological vertices and horizontal bars and scissor braces as topological edges. By inputting the spatial coordinates of the vertices and the cross-sectional mechanical parameters of the edges based on design drawings, the stiffness weights of each edge are determined. The connection constraints of fastener-type or disc-type nodes are also considered, forming a force transmission topology diagram that includes both geometric position and force transmission weights. This topology diagram provides a unified structural framework for the correlation analysis of measured data, enabling the monitoring system to analyze load transfer relationships on a graphical structure, rather than simply relying on independent numerical judgments of each sensor point.

[0038] By deploying sensors on key uprights to collect axial force time-series data, the similarity of axial force fluctuation curves of adjacent and related nodes is calculated, and a correlation matrix reflecting the linkage strength between nodes is constructed and dynamically updated according to a preset time step. Based on the correlation matrix, a first-order neighboring node group is extracted around the target node. The neighboring axial force is weighted and averaged using the linkage strength to obtain a benchmark value of the neighboring axial force. The deviation of the target node's axial force from this benchmark value is used as an anomaly judgment index. Combining the different stress characteristics of the beam rib area and the plate strip area, regional safety thresholds are set separately. When the axial force deviation exceeds the threshold of the region, it is marked as an abnormal node. Thus, relative deviation and regional difference thresholds replace the single absolute value exceeding the limit judgment, improving the rationality of anomaly identification.

[0039] When an abnormal node is detected, an observation time window is set at the moment when the first abnormal node exceeds the limit. The node with the largest axial force deviation within the time window is selected as the starting point. Connections with low linkage strength are filtered out based on the correlation matrix. A recursive search is performed using strong connection edges as paths to form candidate transmission paths. The path with the largest cumulative linkage strength value is selected as the main transmission path. The load is traced back along this path in the opposite direction of load transmission to the projection boundary of the top support layer's grid beam ribs. The end pole of the path is identified as the load disturbance input point. The beam rib segment to which it belongs and the adjacent plate strip area are defined as abnormal load change areas and reported. This achieves the transformation from detecting abnormal points to locating abnormal areas and disturbance input locations at the monitoring level.

[0040] Simultaneously collect vertical deformation time-series data for each node, and use the correlation matrix to calculate the neighborhood deformation benchmark value to obtain the deformation deviation degree of the target node, forming a deviation coefficient with the axial force deviation degree. By pre-setting a reasonable range for the deviation coefficient, when the coefficient is within the range, the node can be considered to be in a normal elastic stress state. When it deviates significantly from the range, it indicates possible structural problems such as decreased upright stiffness, node loosening, foundation settlement, or abnormal constraints. This distinguishes between normal load changes and structural performance degradation, improving the pertinence of anomaly diagnosis.

[0041] A computational model corresponding to the actual formwork is pre-built using finite element method (FEM) software to simulate various typical failure modes. Load transfer paths under failure scenarios are extracted and converted into feature vectors, forming a sample library. During actual monitoring, the main transfer paths obtained through tracing are converted into feature vectors and compared with the sample library for similarity. The real-time identification results are verified and corrected using the corresponding abnormal areas in the sample library, reducing the uncertainty in path identification caused by sensor deployment and on-site disturbances, and improving the reliability of abnormal area location.

[0042] In response to the uneven load distribution and significant transmission and diffusion of abnormal stress in the concrete pouring stage of ultra-high and large-span cold storage grid beam structures, this invention improves the positioning accuracy and early warning reliability of abnormal load areas through continuous domain anomaly diagnosis and source tracing of topological structures. It provides clear technical support for on-site risk identification, construction organization adjustment and emergency response, and has good engineering application prospects. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating the implementation process of the present invention;

[0044] Figure 2 This is a flowchart illustrating the implementation process of the association matrix.

[0045] Figure 3 A flowchart illustrating the implementation process for determining the degree of axial force deviation;

[0046] Figure 4 A logical block diagram for tracing the source of load disturbances;

[0047] Figure 5 Here is the logic block diagram for the joint determination of axial force and deformation;

[0048] Figure 6 This is a time-series monitoring curve of the axial force of each pole during construction;

[0049] Figure 7 This is a schematic diagram for reverse tracing of load disturbances and delineation of abnormal areas. Detailed Implementation

[0050] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0051] Example 1

[0052] This embodiment aims to provide a method for safety monitoring of formwork in ultra-high, large-span cold storage grid beam structures. Addressing issues such as uneven load distribution, transmissibility and diffusion of abnormal stress during concrete pouring in ultra-high, large-span cold storage grid beam structures, and the inability of existing single-point monitoring methods to trace the source of load disturbances, this embodiment achieves precise location and reporting of abnormal load variation areas in the formwork system by establishing a digital force transmission topology map, constructing a load transmission correlation matrix between nodes, and tracing load disturbance input points in reverse along the topology path. This provides on-site management personnel with a basis for quickly identifying potential hazard areas and taking appropriate action.

[0053] The engineering scenario applied in this embodiment is a super-high, large-span cold storage construction project. The cold storage has a floor height of 12.5m, and the roof structure adopts a two-way grid beam arrangement. The main beam cross-section dimensions are 400mm×1200mm, the secondary beam cross-section dimensions are 300mm×900mm, the beam rib spacing is 2400mm, and the slab thickness is 120mm. The formwork system adopts a disc-lock steel pipe scaffold, with the upright spacing being 600mm×600mm in the area below the beam ribs and 900mm×900mm in the non-rib plate area. The total erection height of the scaffold is 11.8m, and the step distance is 1500mm. This formwork system belongs to a sub-project with relatively high risk exceeding a certain scale, and according to relevant regulations, a special construction plan demonstration and full-process safety monitoring are required.

[0054] During concrete pouring, the formwork system bears multiple loads, including the weight of the freshly poured concrete, the weight of the formwork itself, construction live loads, and pumping impact loads. The area below the ribs of the grid beams becomes the primary area of ​​concentrated load due to the large cross-section and heavy weight of the beams, with the axial force of the uprights being significantly higher than in the non-ribbed slab areas. During floor concrete pouring, there is a significant difference in axial force between the middle and edge uprights under the frame beams in the formwork support system. The high-support formwork system experiences complex stresses, making monitoring and control difficult and requiring stricter management. Traditional monitoring methods can only set alarm thresholds for single physical quantities, failing to determine whether an anomaly stems from insufficient load-bearing capacity of the node itself or from load transfer from surrounding areas. Furthermore, as a spatial structure composed of numerous members connected by nodes, the formwork system exhibits significant transmissibility and diffusion of stress states; isolated point monitoring cannot cover the entire process of load anomalies propagating along the topological path.

[0055] like Figure 1 As shown, this embodiment achieves safety monitoring of the formwork system through the following steps: by establishing a digital force transmission topology diagram containing geometric topology information and mechanical transmission weights, and on this basis constructing a load transmission correlation matrix between nodes, the monitoring system no longer relies solely on single-point axial force or displacement exceeding limits, but can comprehensively evaluate the stress state of the entire formwork system.

[0056] S1. Establish a digital force transmission topology diagram.

[0057] This embodiment establishes a digital force transmission topology diagram based on the geometric topological relationships and component connection attributes of the formwork system. Specifically, the uprights with axial load-bearing capacity in the formwork system are designated as topological vertices, and the horizontal bars and scissor braces are designated as topological edges connecting these vertices. The spatial coordinates of the vertices and the cross-sectional mechanical parameters of the topological edges are entered according to the design drawings, and the stiffness weights of each topological edge are obtained based on these parameters. For disc-lock steel pipe supports, the horizontal bars and uprights are connected via disc-lock nodes. The connection stiffness is determined by the construction form of the disc-lock nodes, exhibiting higher rotational stiffness and more stable force transmission performance compared to coupler-type nodes. Based on the construction attributes of the disc-lock nodes, the connection constraints between vertices are determined. By integrating the spatial coordinates, stiffness weights, and connection constraints, a digital force transmission topology diagram containing geometric topological information and mechanical transmission weights is generated.

[0058] Topological graphs are stored in the form of directed graphs in graph theory, denoted as . ,in For the set of vertices (pole). It is a set of topological edges (horizontal bars and scissor braces). This is the set of edge weights (stiffness weights). Each topological edge... stiffness weights Based on the cross-sectional area of ​​the member Elastic modulus and calculation length Calculated using the following formula:

[0059]

[0060] In this embodiment, the stiffness weight This reflects the ease with which loads are transferred along the topological edge. A larger stiffness weight indicates a more significant contribution of the member to load transfer, and higher efficiency in load transfer along the path. By establishing a digital force transmission topology diagram, the members and their connections in the formwork system are transformed into a computable data structure, providing a structured data foundation for subsequent node association analysis and load path tracing. This embodiment uses a disc-lock scaffold as an example, but this method is also applicable to other forms of formwork systems such as coupler-type scaffolds and cup-lock scaffolds. When applied to coupler-type scaffolds, the connection constraint relationships need to be adjusted accordingly based on the semi-rigid characteristics of the coupler nodes.

[0061] S2. Deploy the sensor network and collect axial force timing data.

[0062] A sensor network is deployed at key nodes of the digital force transmission topology to synchronously collect axial force time-series data from each node. The deployment of the sensor network follows these principles: axial force sensors are installed on the tops of the uprights below the rib projection of the grid beam; monitoring nodes are selected at uniform intervals along the topology in non-rib plate areas; and additional sensors are deployed at boundary nodes and geometric abrupt changes in the topology. In this embodiment, vibrating wire axial force sensors are installed on the tops of the uprights below the beam ribs, and sensors are deployed at nodes spaced three uprights apart in non-rib plate areas. All sensors synchronously collect axial force time-series data from each node via a wireless data acquisition module at a sampling frequency of 1Hz, and the data is collected to the on-site monitoring host via LoRa wireless transmission. Data is collected continuously from the start of concrete pouring to the initial curing stage after pouring.

[0063] Figure 6 This paper presents the dynamic evolution curves of axial force at measuring points (N1-N9) within the monitoring window of the concrete layering system. Different colored lines in the figure represent the axial force time-series data of the uprights at different spatial locations. The curve trends show that the axial force at measuring point N5, located directly below the center of the beam rib, is significantly greater than other measuring points, with a peak value of approximately 350 kN and relatively drastic fluctuations. In contrast, the axial force at measuring points located in the slab strip area (such as N1, N2, and N3) has a smaller baseline and relatively gentler changes. The axial force time-series data reflects the real-time stress state of each upright during construction. During concrete pouring, the axial force of the uprights exhibits a phased increase characteristic with the progress of pouring. The synchronous collection of axial force time-series data at each node provides raw data support for subsequent inter-node correlation analysis and axial force deviation calculation.

[0064] S3. Construct the load transfer correlation matrix

[0065] Based on the axial force time-series data of neighboring nodes, the load transfer correlation coefficient between nodes is calculated, and an association matrix representing the degree of force transmission between nodes is constructed. The association matrix combines the measured axial force time-series data with spatial topological relationships, quantifies the linkage strength between nodes, significantly improves the ability to identify load transfer paths and diffusion ranges, and realizes the transformation from isolated monitoring points to the diagnosis of the overall force network.

[0066] In specific implementation, such as Figure 2 As shown, axial force time-series data of each neighboring node within the same monitoring window are extracted, and the similarity of the axial force fluctuation curves of any two nodes is calculated. This similarity is used as the linkage strength between nodes. Let the nodes... and nodes Axial force time series data within the same monitoring window T respectively and If t=1,2,…,n, then the linkage strength between the two is... Calculated based on Pearson correlation coefficient:

[0067]

[0068] In the formula, and They are nodes and nodes The average axial force within the monitoring window. The value range is [−1, 1], where a positive value indicates that the axial force fluctuations of the two nodes change in the same direction, and a negative value indicates that they change in opposite directions. During the load transfer process of the formwork system, the axial force fluctuations of adjacent nodes usually exhibit a characteristic of changing in the same direction. When the load on a certain area increases, the axial force of the uprights in that area rises, and part of the load is transferred to the adjacent uprights through the horizontal bars, causing the axial force of the adjacent uprights to also rise accordingly. The value is primarily positive. The larger the absolute value, the higher the synchronicity of the axial force fluctuations at the two nodes, and the closer the load transfer between them.

[0069] The linkage strength between nodes is normalized, and an affinity matrix R is generated by arranging the topological vertices according to their connection relationships. The affinity matrix R is an N x N symmetric matrix (N is the total number of topological vertices). The specific affinity matrix is ​​shown below:

[0070]

[0071] In the formula, Represents a node and nodes The strength of the linkage between them; when i=j, When node and nodes When there are no topological edges connecting the graphs. When node and nodes When topological edges exist, .

[0072] This correlation matrix integrates the connection relationships in the topology diagram with the measured load transfer intensity. The topology edges represent the physical connections of the members, while the linkage intensity in the correlation matrix reflects the actual load transfer effect on that connection.

[0073] During the construction of high formwork, the layered pouring of concrete leads to a gradual increase in load, and the load transfer path and the linkage relationship between each node also change dynamically. In this embodiment, the linkage strength between nodes is calculated and the correlation matrix is ​​updated according to the preset time step (set to every 5 minutes in this embodiment), so as to realize the dynamic refresh of the correlation matrix to adapt to the characteristics of load distribution changing with time during concrete pouring.

[0074] S4. Calculate the degree of deviation of axial force at nodes.

[0075] like Figure 3 As shown, based on the correlation matrix, a neighborhood node group with a preset topological distance is determined for each node, and the axial force deviation of each node is calculated. In specific implementation, with the target node as the center, all nodes in the first-order neighborhood of the correlation matrix with a topological distance of 0 are extracted to form a neighborhood node group; the topological distance is defined as the minimum number of edges that need to be traversed along the topological edge between two nodes, and the first-order neighborhood is all nodes that are directly connected to the target node through a topological edge.

[0076] Let the target node be Its first-order neighborhood node set is denoted as For any neighboring node in the set There is a direct topological edge connecting it to the target node. Based on the linkage strength values ​​in the correlation matrix R, the weighted average of the axial force monitoring values ​​of the neighboring node group is calculated to obtain the benchmark value of the neighborhood axial force. The calculation formula is as follows:

[0077]

[0078] in, For the target node The neighborhood axial force reference value, For the target node With neighboring nodes The strength of the linkage between them Neighboring nodes The real-time axial force monitoring value is used. The calculation logic of this weighted average is as follows: the higher the linkage strength of a neighboring node, the closer its axial force change is to the target node, and it should be given a higher weight when assessing the normal fluctuation range of the target node's axial force. Based on the linkage strength values ​​in the correlation matrix, the weighted average of the axial force monitoring values ​​of the neighboring node group is calculated to obtain the benchmark value of the neighboring axial force. The higher the linkage strength of a neighboring node, the closer its axial force change is to the target node, and it should be given a higher weight when assessing the normal fluctuation range of the target node's axial force, based on the target node's axial force monitoring value. Compared with the neighboring axial force reference value The absolute value of the difference is used as the degree of axial force deviation. The specific calculation formula is as follows:

[0079]

[0080] The degree of axial force deviation reflects the magnitude of the deviation of the axial force of the target node relative to the weighted average of the axial forces of its neighboring nodes. When an upright experiences abnormal axial force due to insufficient load-bearing capacity, its axial force monitoring value will deviate significantly from the weighted average of the axial forces of its neighbors; similarly, when an upright passively bears additional load due to the transfer of surrounding loads, its axial force deviation will also increase.

[0081] This embodiment takes the target node and its first-order neighboring node group as the object, and uses the linkage strength to perform a weighted average of the neighboring axial force monitoring value to obtain the neighboring axial force benchmark value. The deviation of the target node's axial force from the benchmark value is used as the anomaly judgment index, making the anomaly identification in different stress level areas more accurate.

[0082] S5. Determination of Regionally Differentiated Security Thresholds

[0083] Predefined safety thresholds are used for different regions. When the axial force deviation of a node exceeds the preset safety threshold for the corresponding region, the node is marked as an abnormal node. A first safety threshold is set for nodes located below the rib projection of the grid beam, and a second safety threshold is set for nodes located in the non-rib region of the grid beam, wherein the second safety threshold is less than the first safety threshold.

[0084] This embodiment sets thresholds differently based on the load distribution characteristics of the grid beam structure. Specifically, the area below the beam ribs is the main area of ​​concentrated load, with a large base value of axial force on the uprights. During concrete pouring, the axial force fluctuation amplitude caused by factors such as pumping impact and vibration operations is also relatively large. In contrast, the load in the slab strip area is relatively uniform, and the axial force fluctuation amplitude is smaller. This avoids false alarms caused by normal construction disturbances in the beam rib area, while avoiding missed alarms in the slab strip area due to excessively high thresholds. When the axial force deviation of the target node exceeds the safety threshold of the corresponding area, the abnormal node marking process is triggered.

[0085] S6. Tracing the source of load disturbance and defining abnormal areas

[0086] like Figure 4 As shown, based on the linkage strength of the correlation matrix, starting from the abnormal node, the load is traced back along the topological edge in the opposite direction of load transfer to the projection boundary of the grid beam rib of the top support layer. The end pole of the path on the projection boundary is recorded as the load disturbance input point. The grid beam rib segment to which the load disturbance input point belongs and the adjacent plate strip area are defined as the abnormal load variation area.

[0087] When multiple anomalous nodes are marked, the observation time window is set starting from the first time the first anomalous node exceeds its limit, and the node with the largest axial force deviation within the observation time window is taken as the starting point. During concrete pouring, load disturbances typically begin in a local area and then spread to the surrounding areas along the topological path. The first time the first anomalous node exceeds its limit marks the start time of the load disturbance, while the node with the largest axial force deviation within the observation time window represents the center of the area where the disturbance has the most severe impact.

[0088] During path tracing, weak connections are filtered out based on the linkage strength of the correlation matrix. A recursive search is then performed from the abnormal node along the strong connections towards the topological boundary. Weak connections are defined as topological edges with linkage strength below a preset threshold; the members corresponding to these edges contribute little to the actual load transfer and are not considered part of the main transfer path. The sequence of nodes traversed along the path is recorded as candidate transfer paths. The cumulative value of linkage strength in each candidate transfer path is calculated, and the candidate transfer path with the highest cumulative value is determined as the main transfer path.

[0089] Tracing back against the load transfer direction, that is, searching in the opposite direction of the main transfer path from the abnormal node until reaching the projection boundary of the grid beam ribs of the top support layer. The grid beam ribs are the direct area of ​​concentrated load action. The concrete self-weight of the beam ribs acts directly on the vertical poles below through the formwork and top support, which is the entry point for the load into the formwork system. The vertical poles where the path terminates on the projection boundary are the load disturbance input points. The load changes borne by these vertical poles are directly caused by the load input of the corresponding section of the grid beam ribs above. The grid beam rib section to which the load disturbance input point belongs and the adjacent slab area are defined as the abnormal load variation area.

[0090] like Figure 7 The demonstration shows that when the system detects an abnormal node's axial force deviation exceeding the limit, it recursively searches upwards along the strong connection edge (red solid line) based on the linkage strength in the correlation matrix. The dark red curve path in the figure shows the specific reverse tracing trajectory, which runs against the load transfer direction and ultimately locks onto the load disturbance input point (orange node) on the projection boundary of the top support beam rib. Based on this, the system automatically defines the grid beam rib segment to which the input point belongs and the adjacent plate strip area (purple shaded area) as the abnormal load change area, transforming the traditional passive single-point alarm into proactive source diagnosis. On-site management personnel can quickly locate the root cause of the problem and take targeted measures based on the load transfer path and defined abnormal area shown in the figure.

[0091] This embodiment relies on the linkage strength of the correlation matrix. After identifying abnormal nodes, it recursively searches for candidate transmission paths along the strong connection edges, and uses the path with the largest cumulative linkage strength value as the main force transmission path. It traces back to the projection boundary of the top support layer's grid beam ribs in the opposite direction of load transmission, accurately locating the load disturbance input point. The beam rib segment to which the load disturbance input point belongs and its adjacent plate strip area are automatically defined as abnormal load change areas, and information such as location coordinates, range, and main transmission path are reported. This provides on-site management personnel with intuitive and clear evidence for quickly locating potential hazard areas and finding their causes (such as local over-grouting, pumping impact overload, excessive vibration, etc.).

[0092] S7. Reporting of Abnormal Load Variation Areas

[0093] Information on identified abnormal load variations is reported to the monitoring platform. This report includes the area's location coordinates, extent (beam rib segment number and adjacent slab range), time of occurrence, extreme values ​​of axial force deviation at the abnormal nodes, and the main transmission path. Based on this information, on-site management personnel can quickly locate the potential hazard area, check for any abnormalities in the concrete pouring operation (such as localized over-pouring, pump pipe impact overload, excessive vibration, etc.), and take targeted corrective measures.

[0094] This embodiment constructs a digital force transmission map that integrates geometric topology and mechanical weights, mapping discrete sensor data into a high-dimensional mechanical network space. It then calculates the Pearson correlation coefficient of the axial force time series of neighboring nodes in real time to construct a dynamic correlation matrix, quantifying the load transfer tightness between nodes. Furthermore, based on the linkage strength, it performs a weighted average of the neighboring axial forces to establish an adaptive axial force benchmark value. By calculating the deviation of the target node from this benchmark value, it accurately separates normal construction disturbances from actual structural anomalies. On this basis, it traces the load transfer path backward along the strong connection edge to the projection boundary of the top support beam rib, locking the load disturbance input point and defining the abnormal region. This embodiment breaks through the limitations of traditional single-point threshold alarms, effectively solving the problem of false alarms and missed alarms caused by the force transmission of high formwork. It achieves a qualitative change from single-point abnormal alarms to regional source diagnosis, significantly improving the predictability and handling efficiency of safety management in the construction of ultra-high and large-span cold storage. It makes full use of the spatial topological relationship of the formwork system and the correlation characteristics of measured axial force data, providing an effective technical means for the safety management of the formwork construction of the grid beam structure of ultra-high and large-span cold storage.

[0095] Example 2

[0096] Based on Example 1, this embodiment further introduces vertical deformation monitoring data. By analyzing the ratio of axial force deviation to deformation deviation, the working status of nodes can be classified and judged, thereby improving the accuracy of anomaly diagnosis.

[0097] During the service life of the formwork system, an abnormal increase in the axial force of the uprights may originate from two different physical mechanisms: First, normal load transfer under elastic stress. When the construction load on a certain area increases, the axial force of the uprights in that area rises, accompanied by corresponding elastic compressive deformation; axial force and deformation are positively correlated. Second, deterioration of nodes or members (such as upright bending, loosening of nodes, foundation settlement, etc.). In this case, the axial force distribution redistributes, and the deformation characteristics become abnormal, with the relationship between axial force and deformation deviating from the normal range. Distinguishing between these two mechanisms is crucial for taking appropriate on-site measures: the former is a normal load response, requiring only attention to whether the load exceeds the design value; the latter is a structural anomaly, necessitating immediate work stoppage and investigation.

[0098] This embodiment, based on the synchronous acquisition of axial force time-series data at each node, further acquires vertical deformation time-series data at each node. A laser displacement gauge or a wire-type displacement sensor is installed on the upper part of the pole, acquiring data synchronously with the axial force sensor. Based on the linkage strength value in the correlation matrix, a weighted average is calculated for the vertical deformation time-series data of the neighboring node group to obtain the neighborhood deformation reference value. The absolute value of the difference between the vertical deformation time-series data of the target node and the neighborhood deformation reference value is calculated and defined as the degree of deformation deviation. The ratio of the axial force deviation degree to the deformation deviation degree is calculated as the deviation coefficient for that node, its physical meaning being the axial force deviation corresponding to a unit deformation deviation.

[0099] Under normal stress conditions, the ratio of axial force deviation to deformation deviation should be consistent with the ratio of axial stiffness. When nodes or members deteriorate, their axial equivalent stiffness decreases, resulting in a larger deformation change under the same axial force change, leading to a smaller deviation coefficient value. Conversely, when the deformation of a node is restricted due to factors such as enhanced constraints, the deviation coefficient value is larger.

[0100] like Figure 5 As shown, this embodiment sets a first preset interval, and determines the upper and lower limits of the interval based on the theoretical value of the axial stiffness of the upright and engineering experience. If the deviation coefficient is within the first preset interval, it is determined that the node is in a normal elastic stress state, and the load disturbance source tracing and abnormal area definition are continued according to the steps of Embodiment 1. If the deviation coefficient is less than the lower limit of the interval, it indicates that the axial equivalent stiffness of the node or member is abnormally low. Possible reasons include bending deformation of the upright, loosening of the disc-lock node leading to a reduction in effective connection stiffness, uneven settlement of the foundation, and other structural deterioration phenomena. At this time, a structural abnormality alarm containing location information is reported, and work is immediately stopped for investigation. If the deviation coefficient is greater than the upper limit of the interval, it indicates that the node deformation is restricted. Possible reasons include the node being jammed by additional constraints, abnormal sensor installation, etc. In this case, an abnormality alarm is also reported and manual verification is recommended.

[0101] This embodiment introduces deformation monitoring data and performs joint analysis with axial force data to classify and distinguish the working state of nodes, effectively differentiating between normal load transfer response and abnormal response caused by structural deterioration, thereby improving the accuracy of anomaly diagnosis and the pertinence of treatment recommendations.

[0102] Example 3

[0103] Based on Example 1, this embodiment further introduces a load transfer path sample library matching mechanism based on finite element simulation. By matching the main transfer path determined by real-time monitoring with the path feature vectors in the simulation sample library, the definition results of abnormal load variation areas are verified and corrected, thereby improving the reliability of abnormal area definition.

[0104] In practical engineering, due to factors such as limited sensor deployment density and on-site construction disturbances, the load transfer path traced based on measured data may have certain uncertainties. Therefore, this embodiment constructs a sample library of formwork load transfer paths under different failure modes using finite element simulation software before construction. Specifically, finite element analysis software such as ANSYS or ABAQUS is used to establish a refined finite element model based on the design drawings of the formwork system. The model includes all components such as uprights, horizontal bars, scissor braces, and disc-lock nodes. Uprights are simulated using beam elements, horizontal bars using truss elements, and the disc-lock nodes have their rotational stiffness set according to their structural characteristics.

[0105] Multiple failure modes were simulated in the finite element model, including single-pillar insufficient bearing capacity failure, local overload failure, node connection failure, and uneven foundation settlement failure. For each failure mode, corresponding load conditions or boundary conditions were applied to the model, and nonlinear buckling analysis or explicit dynamic analysis was performed to extract the load transfer path from the disturbance source to each anomalous node. Each transfer path was converted into a feature vector form, where the dimension of the feature vector is the number of each topological edge in the topological graph, and the vector elements are the presence or absence of the edge in the path or the weight coefficient of the edge in the path. A mapping relationship between the path feature vectors and anomalous regions was established to form a load transfer path sample library.

[0106] During real-time monitoring, after determining the main transmission path according to the steps in Example 1, the path is converted into a feature vector of the same format and similarity matching is performed in the sample library. Cosine similarity is used to calculate the similarity. All paths in the sample library are traversed to find the sample paths with the highest similarity to the real-time path (in this example, the top 3 with the highest similarity are selected). The abnormal load variation areas defined in Example 1 are verified and corrected based on the abnormal areas corresponding to these sample paths. If the abnormal areas corresponding to multiple highly similar sample paths are consistent with the real-time definition results, the reliability of the definition results is verified; if there are discrepancies, the abnormal areas are corrected based on the sample library matching results, and on-site personnel are prompted to conduct focused checks on these areas.

[0107] This embodiment introduces a finite element simulation sample library matching mechanism to cross-verify the measured data tracing results with the theoretical simulation results. This effectively reduces the uncertainty of path tracing caused by factors such as limited sensor deployment density or on-site disturbances, and improves the reliability and accuracy of defining abnormal load variation areas.

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

Claims

1. A method for safety monitoring of the formwork of a grid beam structure for ultra-high, large-span cold storage facilities, characterized in that, Includes the following steps: S1. Based on the geometric topological relationship and component connection properties of the formwork system, establish a digital force transmission topology diagram; S2, Deploy a sensor network at the key nodes of the topology to synchronously collect axial force time-series data of each node; S3, based on the axial force time series data of neighboring nodes, calculate the load transfer correlation coefficient between nodes and construct an association matrix representing the degree of force transfer between nodes; Constructing the association matrix includes: Extract the axial force time series data of each neighboring node within the same monitoring window, calculate the similarity of the axial force fluctuation curves of any two nodes, and use the similarity as the linkage strength between nodes; The linkage strength between each node is normalized and the association matrix is ​​generated according to the connection relationship of the topological vertices. S4. Based on the correlation matrix, determine the neighborhood node group with a preset topological distance centered on each node, and calculate the axial force deviation of each node; with the target node as the center, extract all nodes in the first-order neighborhood of the correlation matrix to form a neighborhood node group. Based on the linkage strength values ​​in the correlation matrix, the weighted average value of the axial force monitoring values ​​of the neighboring node group is calculated to obtain the neighborhood axial force benchmark value; the absolute value of the difference between the axial force monitoring value of the target node and the neighborhood axial force benchmark value is used as the degree of axial force deviation. S5, predefine safety thresholds for different regions. When the axial force deviation of a node exceeds the preset safety threshold of the corresponding region, the node is marked as an abnormal node. S6. Based on the linkage strength of the correlation matrix, starting from the abnormal node, trace back along the topological edge against the load transfer direction to the projection boundary of the grid beam rib of the top support layer, and record the terminal column of the path on the projection boundary as the load disturbance input point; define the grid beam rib segment and adjacent plate area to which the load disturbance input point belongs as the abnormal load change area. S7, report areas with abnormal load changes.

2. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to claim 1, characterized in that, Constructing a digital force transmission topology map includes: The uprights with axial force bearing capacity in the formwork system are taken as topological vertices, and the horizontal bars and scissor braces are taken as topological edges connecting the vertices. The spatial coordinates of the vertices and the cross-sectional mechanical parameters of the topological edges are entered according to the design drawings, and the stiffness weights of each topological edge are obtained according to the cross-sectional mechanical parameters. Based on the node construction attributes of fastener-type or disc-type nodes, determine the connection constraint relationships between vertices. By integrating the spatial coordinates, stiffness weights, and connection constraints, a digital force transmission topology diagram containing geometric topology information and mechanical transmission weights is generated.

3. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to claim 1, characterized in that, The linkage strength between nodes is calculated and the correlation matrix is ​​updated according to the preset time step.

4. The method for monitoring the safety of the formwork of the ultra-high, large-span cold storage grid beam structure according to claim 1, characterized in that, For nodes located below the projection of the ribs of the grid beam, a first safety threshold is set; For nodes located in the non-ribbed area of ​​the grid beam, a second safety threshold is set, wherein the second safety threshold is less than the first safety threshold; When the deviation of the axial force exceeds the safety threshold of the corresponding area, the marking process of the abnormal node is triggered.

5. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to claim 1, characterized in that, When there are multiple marked abnormal nodes, the first time the first abnormal node exceeds the limit is taken as the starting point, and the observation time window is set; the node with the largest axial force deviation within the observation time window is taken as the starting point.

6. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to claim 1, characterized in that, Weak connections are filtered out based on the linkage strength of the correlation matrix. Starting from the abnormal node, a recursive search is performed along the strong connection edge towards the topological boundary, and the sequence of nodes passed along the way is recorded as a candidate transmission path. Calculate the cumulative value of linkage intensity in each candidate transmission path, and determine the candidate transmission path with the largest cumulative value as the main transmission path.

7. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to claim 1, characterized in that, Simultaneously collect the vertical deformation time series data of each node; based on the linkage intensity value in the correlation matrix, calculate the weighted average value of the vertical deformation time series data of the neighboring node group to obtain the neighborhood deformation benchmark value; The absolute value of the difference between the vertical deformation time series data of the target node and the deformation reference value of the neighborhood is calculated and defined as the degree of deformation deviation. Calculate the ratio of the axial force deviation to the deformation deviation, and use it as the deviation coefficient for that node; If the deviation coefficient is within the first preset range, the node is determined to be normal, and S6 is performed. If the deviation coefficient is greater than the upper limit of the first preset interval, the node is determined to be degraded, and a structural anomaly alarm containing location information is reported.

8. The method for monitoring the formwork safety of a super-high, large-span cold storage grid beam structure according to any one of claims 1-7, characterized in that, Based on finite element simulation software, a sample library of formwork load transfer paths under different failure modes is constructed, and a mapping relationship between load transfer path feature vectors and abnormal regions is established. The main transfer paths determined in real time are converted into feature vectors, and similarity matching is performed in the sample library to verify and correct the definition results of the abnormal load variation regions.

Images