Lightweight steel arch service state big data evaluation and decision system

Abstract

The present application relates to the technical field of structural engineering monitoring, and discloses a big data evaluation and decision system for the service state of a lightweight steel arch, comprising a data acquisition module, a data preprocessing module, a dynamic response coupling graph construction module, a topological spectrum decomposition module, a topological invariant extraction module, a service state feature construction module, a disturbance decoupling module, and a state evaluation and decision module. The system acquires structural response data from multiple monitoring nodes of the lightweight steel arch, constructs a dynamic response coupling graph between the monitoring nodes, extracts topological spectrum features and topological invariant abstract features of the structural response, forms structural service state features, and generates service state parameters according to the time evolution process of the structural intrinsic features. The present application extracts topological spectrum features and topological invariant abstract features of the structural response by combining the topological spectrum decomposition module and the topological invariant extraction module, so that the service state of the lightweight steel arch is characterized by node correlation structure.

Description

Technical Field

[0001] This invention relates to the field of structural engineering monitoring technology, and in particular to a big data assessment and decision-making system for the service status of lightweight steel arch frames. Background Technology

[0002] Lightweight steel arch frames are widely used in tunnel engineering, underground engineering support structures, and other load-bearing structures. These structures typically employ a multi-point monitoring system to monitor structural strain, displacement, and stress conditions in real time.

[0003] In existing technologies, the service status assessment of lightweight steel arch frames is typically based on the response values ​​of individual monitoring nodes. Common methods include threshold judgment, statistical analysis, or trend comparison of monitoring node data, and determining the structural status based on whether the response of a single node exceeds a preset limit. Some solutions use simple statistical summaries of data from multiple monitoring nodes, such as calculating the average or standard deviation, to describe the overall structural status. Additionally, there are technical solutions that use structural numerical models to perform back-calculation analysis on the monitoring data.

[0004] However, the inventors of this application discovered in the process of implementing the technical solution of this application that the prior art has at least the following technical problems: the existing service status assessment methods based on single node response values ​​or simple statistics cannot model the overall correlation structural changes between multiple monitoring nodes of lightweight steel arch frames, resulting in the expression of the overall service status of the structure lacking structural information at the level of coupling relationship between nodes. Summary of the Invention

[0005] To overcome the above shortcomings, this invention provides a big data assessment and decision-making system for the service status of lightweight steel arch frames. It aims to improve the problem that existing technologies cannot model the overall structural changes in the interconnected structure between multiple monitoring nodes of lightweight steel arch frames, resulting in a lack of structural information at the level of inter-node coupling relationships in the expression of the overall service status of the structure.

[0006] This invention provides the following technical solution: a big data assessment and decision-making system for the service status of lightweight steel arch frames, comprising: The data acquisition module collects structural response data generated by multiple monitoring nodes arranged along the lightweight steel arch frame structure, wherein the monitoring nodes correspond to different structural positions of the lightweight steel arch frame. The data preprocessing module standardizes the structural response data and constructs a sliding time window data sequence in chronological order. The dynamic response coupling diagram construction module calculates the response correlation between monitoring nodes based on the spatial arrangement of monitoring nodes in the lightweight steel arch frame structure and the sliding time window data sequence, and forms a dynamic response coupling diagram characterizing the response correlation state of the lightweight steel arch frame structure. The topology spectrum decomposition module constructs a structural response topology model based on the dynamic response coupling diagram and obtains the corresponding structural response topology spectrum feature data. The topological invariant extraction module obtains topological invariant summary feature data reflecting the connection morphology changes of the lightweight steel arch frame structure based on the multi-threshold evolution process of the dynamic response coupling graph. The service status feature construction module constructs service status features of the lightweight steel arch frame structure based on the structural response topology spectrum features and the topology invariant summary features. The disturbance decoupling module combines external disturbance data to perform topological layered projection processing on the service state characteristics of the lightweight steel arch frame structure and obtains the intrinsic feature data of the structure. The condition assessment and decision-making module obtains the service condition parameters of the lightweight steel arch frame based on the time evolution process of the intrinsic characteristics of the structure.

[0007] Preferably, in the dynamic response coupling graph construction module, the step of forming a dynamic response coupling graph characterizing the response correlation state of the lightweight steel arch frame structure includes: The structural response data of each monitoring node within the sliding time window are synchronized in time. Calculate the time delay correlation between any two monitoring nodes; Calculate the rate-of-change relationship between the two monitoring nodes; The coupling weights between monitoring nodes are determined based on the aforementioned time delay correlation and rate of change synergy. Based on the coupling weights, the connection relationships between monitoring nodes are constructed, and a dynamic response coupling diagram is formed.

[0008] Preferably, the step of constructing the connection relationship between monitoring nodes and generating a dynamic response coupling graph based on the coupling weights includes: Each monitoring node is used as a graph node; Establish weighted connections between nodes according to their coupling weights; The weighted connectivity is updated in time order to form a dynamic response coupling graph sequence.

[0009] Preferably, in the topology spectral decomposition module, the step of constructing a structural response topology model and generating corresponding structural response topology spectral features from the dynamic response coupling graph includes: Establish an adjacency matrix based on the dynamic response coupling graph; Construct the degree matrix based on the adjacency matrix; Generate the graph Laplacian matrix based on the adjacency matrix and degree matrix; Perform eigenvalue decomposition on the graph Laplacian matrix; Based on the decomposition results, the topological spectrum features of the structural response are extracted.

[0010] Preferably, in the topological invariant extraction module, the step of generating a summary of topological invariant features reflecting the connection morphology changes of the lightweight steel arch frame structure through the multi-threshold evolution process of the dynamic response coupling graph includes: The dynamic response coupling diagram is filtered step by step according to multiple preset coupling strength thresholds; Generate corresponding threshold sub-graphs based on each coupling strength threshold; Construct a simple complex structure based on the threshold subgraph; Perform topological statistical processing on the simple complex structure; Generate topological invariant summary features based on statistical results.

[0011] Preferably, the step of performing topological statistical processing on the simple complex structure includes: Count the number of connected components in the threshold subgraph; Count the number of closed connection structures in the threshold subgraph; The statistical results are accumulated according to the order of threshold changes.

[0012] Preferably, in the service status feature construction module, the step of constructing the service status features of the lightweight steel arch structure from the structural response topology spectrum features and the topology invariant summary features includes: The topological spectrum features of the structural response are arranged in a unified dimension. The topological invariant summary features are arranged in a unified dimension; The two types of features are combined in a preset order; Generate service status feature vectors for the corresponding time series.

[0013] Preferably, in the disturbance decoupling module, the step of performing topological layered projection processing on the service state characteristics of the lightweight steel arch frame structure based on external disturbance data and generating intrinsic structural features includes: Obtain time series data of external disturbances; Establish the correspondence between external disturbance data and service status feature vectors; Construct a perturbation mapping model based on the correspondence; The service state feature vector is decomposed by projection based on the aforementioned disturbance mapping model; The structural intrinsic features are generated based on the decomposition results.

[0014] Preferably, the step of projecting the service state feature vectors according to the disturbance mapping model includes: Calculate the directional components of the external disturbance; Calculate the directional components excluding external disturbances; The non-external disturbance direction components are retained as structural intrinsic features.

[0015] Preferably, in the condition assessment and decision-making module, the step of generating service condition parameters for the lightweight steel arch frame through the time evolution process of the structural intrinsic features includes: Constructing time-series trajectories of structural intrinsic features; Calculate the relationship between changes in the intrinsic structural features at adjacent time points; The state classification is determined based on the aforementioned change relationship; Output the corresponding service status parameters.

[0016] The present invention has the following beneficial effects: 1. This invention establishes a weighted correlation between monitoring nodes through a dynamic response coupling graph construction module, and extracts the topological spectrum features and topological invariant summary features of the structural response by combining a topological spectrum decomposition module and a topological invariant extraction module, so that the service status of the lightweight steel arch frame can be characterized by the node-related structure.

[0017] 2. This invention uses multi-threshold screening of dynamic response coupling graphs and constructs simple complex structures to statistically analyze the number of connected components and closed connection structures to form topological invariant summary features, enabling structural connection morphology changes to be expressed with unified statistical parameters.

[0018] 3. This invention performs topological layered projection processing on the service state characteristics through a disturbance decoupling module, separates the external disturbance direction components from the structural response components, and generates service state parameters based on the intrinsic characteristics of the structure, thereby achieving the ability to reflect the structural response changes of the lightweight steel arch frame itself. Attached Figure Description

[0019] Figure 1 This is an architecture diagram of a big data assessment and decision-making system for the service status of lightweight steel arch frames proposed in this invention. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Reference Figure 1 This invention provides a big data assessment and decision-making system for the service status of lightweight steel arch frames, comprising: The data acquisition module collects structural response data generated by multiple monitoring nodes deployed along the lightweight steel arch frame structure. The monitoring nodes correspond to different structural positions of the lightweight steel arch frame. The data preprocessing module standardizes the structural response data and constructs a sliding time window data sequence in chronological order. The dynamic response coupling diagram construction module calculates the response correlation between monitoring nodes based on the spatial layout of monitoring nodes in the lightweight steel arch frame structure and the sliding time window data sequence, and forms a dynamic response coupling diagram that characterizes the response correlation state of the lightweight steel arch frame structure. The topology spectral decomposition module constructs a structural response topology model based on the dynamic response coupling graph and obtains the corresponding structural response topology spectral feature data. The topological invariant extraction module obtains topological invariant summary feature data reflecting the connection morphology changes of lightweight steel arch frame structures based on the multi-threshold evolution process of the dynamic response coupling graph. The service status feature construction module constructs service status features of lightweight steel arch frame structures based on structural response topology spectrum features and topology invariant summary features. The disturbance decoupling module combines external disturbance data to perform topological layered projection processing on the service status characteristics of the lightweight steel arch frame structure and obtain the intrinsic feature data of the structure. The condition assessment and decision-making module obtains the service condition parameters of the lightweight steel arch frame based on the time evolution process of the structural intrinsic characteristics.

[0022] Specifically, multiple monitoring nodes are deployed along the structural direction of the lightweight steel arch frame. Each monitoring node is installed at different structural locations on the arch frame, including the arch foot area, arch waist area, and arch crown area. The monitoring nodes are connected to a data acquisition module via a communication interface. The data acquisition module acquires structural response data generated by each monitoring node according to a preset sampling period and forms a continuous data sequence in chronological order of sampling time. A data preprocessing module receives the structural response data and performs unified processing on the data from each monitoring node within the sliding time window. The processing includes amplitude normalization and scale unification of the structural response data to ensure that the data collected by different monitoring nodes are at a consistent data scale, and continuously updates the sliding time window data sequence in chronological order.

[0023] The dynamic response coupling graph construction module performs inter-node correlation calculations on the structural response data within a sliding time window based on the spatial layout of monitoring nodes in the lightweight steel arch structure. This module calculates the synchronization relationship and trend consistency of response changes between any two monitoring nodes, and determines the coupling weights between the monitoring nodes based on the calculation results. Using the monitoring nodes as graph nodes and the coupling weights as the connection relationships between nodes, a dynamic response coupling graph reflecting the structural response correlation state of each monitoring node is constructed and updated over time to form a dynamic response coupling graph sequence. The topology spectrum decomposition module establishes a corresponding graph structure model based on the dynamic response coupling graph. By matrix-representing the node connection relationships in the graph structure, calculating the node connection degree information, and performing feature decomposition on the graph structure, it obtains structural response topology spectrum feature data describing the overall structural response correlation characteristics.

[0024] The topology invariant extraction module sets multiple coupling strength thresholds based on the dynamic response coupling graph, performing step-by-step filtering to generate multiple threshold subgraphs. A simple complex structure is constructed based on the node connectivity in each threshold subgraph, and the number of connected components and closed connections within the simple complex structure is statistically analyzed. The statistical results are accumulated and calculated according to the threshold change order to obtain the topology invariant summary feature data. The service status feature construction module arranges the structural response topology spectrum feature data and the topology invariant summary feature data in a unified dimension and combines them according to a preset order to form the service status features of the lightweight steel arch structure for the corresponding time series.

[0025] The disturbance decoupling module acquires external disturbance data related to the operating environment of the lightweight steel arch frame, including ambient temperature changes, surrounding rock deformation data, and external load changes, and establishes a correspondence between the external disturbance data and service status characteristics. Based on this correspondence, the service status characteristics are subjected to projection decomposition processing to separate the external disturbance-related components from the non-external disturbance components, and the non-external disturbance components are used as the intrinsic structural characteristic data. The status assessment and decision module constructs a continuous time series based on the intrinsic structural characteristic data, calculates the changing relationship between the intrinsic structural characteristics at adjacent time points, and generates the service status parameters of the lightweight steel arch frame at the corresponding time point based on the changing results.

[0026] Furthermore, in the dynamic response coupling graph construction module, the steps for forming a dynamic response coupling graph characterizing the response correlation state of the lightweight steel arch frame structure include: The structural response data of each monitoring node within the sliding time window are synchronized in time. Calculate the time delay correlation between any two monitoring nodes; Calculate the rate-of-change relationship between two monitoring nodes; The coupling weights between monitoring nodes are determined based on the correlation between time delay and the synergistic relationship between the rate of change. The connection relationships between monitoring nodes are constructed based on the coupling weights, and a dynamic response coupling diagram is formed.

[0027] The steps for constructing the connection relationships between monitoring nodes and generating a dynamic response coupling graph based on coupling weights include: Each monitoring node is used as a graph node; Establish weighted connections between nodes according to their coupling weights; The weighted connectivity relationships are updated in chronological order to form a dynamic response coupling graph sequence.

[0028] Specifically, during the operation of the lightweight steel arch structure, each monitoring node collects structural response data according to a unified sampling period. The sliding time window data sequence output by the data preprocessing module contains structural response values ​​at multiple consecutive sampling times. The dynamic response coupling graph construction module first sorts the data from all monitoring nodes within the same sliding time window according to the sampling time, so that different monitoring nodes correspond to the same structural response time under the same time index, thereby forming a time-synchronized data matrix.

[0029] Suppose there are N monitoring nodes deployed on the lightweight steel arch frame, and the standardized structural response data of the i-th monitoring node at the k-th sampling time is denoted as . Where i represents the monitoring node number. This represents the sequence number of the sampling time. After completing the time synchronization arrangement, the time delay correlation is calculated for the structural response sequences between any two monitoring nodes. The time delay correlation is used to characterize the consistency of the structural response changes of the two monitoring nodes under different time offset conditions, and its calculation form is as follows: ; in, This represents the time delay correlation coefficient between monitoring node i and monitoring node j; Indicates the amount of time delay; This indicates the maximum allowable time delay range; W represents the length of the sliding time window.

[0030] The maximum correlation under different time delay conditions is obtained through the above calculations, which is used to describe the time lag correlation characteristics generated during the propagation of structural response.

[0031] After obtaining the time delay correlation, the rate of change coordination relationship between the two monitoring nodes is further calculated. This coordination relationship is determined by comparing the consistency of the structural response change amplitude at adjacent sampling times. First, the rate of change of the structural response at each monitoring node is calculated based on the time series. Then, the rate of change sequences of the two monitoring nodes are synchronously statistically processed to obtain the degree of consistency in their change trends. This rate of change coordination relationship is denoted as... It is used to reflect the consistency of the direction and rate of change of structural response.

[0032] After obtaining the correlation between time delay and the co-relationship between the rate of change, the dynamic response coupling graph construction module fuses the two types of correlation quantities to determine the coupling weight between monitoring nodes. The calculation form is as follows: ; in, This represents the coupling weight between monitoring node i and monitoring node j; , which is a weighting coefficient used to adjust the proportion of the time delay correlation and the rate of change synergy in the coupled calculation; This represents the hyperbolic tangent function, used to perform amplitude constraint processing on correlated values.

[0033] When constructing node connections based on coupling weights, each monitoring node is treated as a node in the graph structure, and the coupling weight is used as the connection strength between nodes. When the coupling weight is greater than a preset connection threshold, a weighted connection is established between the corresponding nodes, thus forming a weighted adjacency structure.

[0034] During the time update process, the system continuously updates the structural response data according to the sliding time window sequence and recalculates the coupling weights between nodes. The updated weighted connection relationships replace the connection relationships at the previous moment, thus forming a dynamic response coupling graph sequence that changes over time. The dynamic response coupling graph sequence is used to reflect the evolution of the response association state of the lightweight steel arch frame structure under continuous time conditions.

[0035] Through the above steps, the original structural response data is transformed into an associated network structure between monitoring nodes, so that the response relationship between different structural locations of the lightweight steel arch frame can be expressed in the form of a graph structure. This provides a unified data structure foundation for subsequent topological spectral decomposition and topological invariant extraction, thereby realizing continuous modeling of the structural response associated state.

[0036] Furthermore, in the topology spectral decomposition module, the steps of constructing a structural response topology model from the dynamic response coupling graph and generating the corresponding structural response topology spectral features include: Establish an adjacency matrix based on the dynamic response coupling graph; Construct the degree matrix based on the adjacency matrix; Generate the graph Laplacian matrix based on the adjacency matrix and degree matrix; Perform eigenvalue decomposition on the graph Laplacian matrix; Based on the decomposition results, the topological spectrum features of the structural response are extracted.

[0037] Specifically, after the dynamic response coupling graph is constructed, the system obtains a weighted graph structure within a certain time window. Assuming N monitoring nodes are deployed on the lightweight steel arch frame, an N×N adjacency matrix is ​​constructed to represent the coupling relationship between any two monitoring nodes.

[0038] The adjacency matrix is ​​denoted as A, and its elements are... This represents the coupling weight between monitoring node i and monitoring node j. When there is a connection between the two nodes, the matrix elements take the corresponding coupling weight; when there is no connection, the matrix elements take zero. The adjacency matrix is ​​a symmetric matrix with zero elements on the diagonal.

[0039] After obtaining the adjacency matrix, construct the degree matrix. The degree matrix is ​​denoted by D, which is a diagonal matrix, and its i-th diagonal element... Let represent the node degree of monitoring node i. The node degree is the sum of the coupling weights of this node with all other nodes, and its calculation form is: ; in, This represents the degree value of the i-th node; Let L represent the element in the i-th row and j-th column of the adjacency matrix; N represents the total number of monitored nodes. After constructing the adjacency matrix and degree matrix, a graph Laplacian matrix is ​​generated. The graph Laplacian matrix is ​​denoted as L, and its definition is: ;in, Let represent the graph Laplacian matrix; D represents the degree matrix; and A represents the adjacency matrix.

[0040] The graph Laplacian matrix is ​​used to characterize the structural connectivity of the entire dynamic response coupled graph. This matrix is ​​a symmetric positive semi-definite matrix, and its eigenvalues ​​are directly related to the overall connectivity of the graph structure. After obtaining the graph Laplacian matrix, it undergoes eigenvalue decomposition. The eigenvalue decomposition process satisfies the following relationship: ;in, This represents the m-th eigenvalue; This represents the corresponding eigenvector; m = 1, 2, ..., N.

[0041] Eigenvalue decomposition yields a sequence of eigenvalues ​​ordered by size. Since the graph Laplacian matrix is ​​a positive semi-definite matrix, its smallest eigenvalue is zero, reflecting the connectivity of the graph structure. The remaining eigenvalues ​​reflect the overall coupling strength distribution and structural association patterns between nodes. When extracting the topological spectrum features of the structural response, a predetermined number of non-zero eigenvalues ​​are selected to form the feature vector. Let the first M non-zero eigenvalues ​​be selected as the topological spectrum features; then the structural response topological spectrum features are expressed as: ; Where S represents the structural response topological spectrum eigenvector; represents the second smallest eigenvalue; M represents the number of features selected.

[0042] To ensure comparability across different time windows, the feature values ​​are arranged in ascending order, and the feature dimensions are kept consistent. The structural response topological spectrum feature data serves as a component in the subsequent construction of service status features.

[0043] Through the above steps, the dynamic response coupling diagram is transformed into quantifiable spectral feature data, and the overall correlation state of the structure is transformed from a graph structure form to a numerical feature form. This enables a unified representation of the response correlation mode of the lightweight steel arch frame structure, providing basic data for subsequent topological invariant extraction and structural intrinsic feature calculation.

[0044] Furthermore, in the topological invariant extraction module, the steps for generating a summary of topological invariant features reflecting the connection morphology changes of the lightweight steel arch frame structure through the multi-threshold evolution process of the dynamic response coupling graph include: The dynamic response coupling diagram is filtered step by step according to multiple preset coupling strength thresholds; Generate corresponding threshold sub-graphs based on each coupling strength threshold; Construct a simple complex structure based on the threshold subgraph; Topological statistical processing of simple complex structures; Generate topological invariant summary features based on statistical results.

[0045] The steps for performing topological statistical processing on simple complex structures include: Count the number of connected components in the threshold subgraph; Count the number of closed connection structures in the threshold subgraph; The statistical results are accumulated according to the order of threshold changes.

[0046] Specifically, after the dynamic response coupling graph is constructed, the system obtains a weighted network structure reflecting the coupling relationships between monitoring nodes. Since different coupling strengths correspond to different levels of structural association states, the topology invariant extraction module performs multi-threshold filtering on the dynamic response coupling graph to form a hierarchical evolution process of structural connection relationships.

[0047] Let A be the adjacency matrix of the dynamic response coupling graph, where the matrix elements are... This represents the coupling weight between monitoring node i and monitoring node j. The system pre-sets a sequence of coupling strength thresholds arranged in a regular order of magnitude: ;in, Let represent the h-th coupling strength threshold; H represents the number of thresholds. For each threshold, the adjacency matrix is ​​filtered to generate a corresponding threshold subgraph. The threshold filtering rules are as follows: ; in, Indicates at the threshold The subgraph connection relationships obtained under the given conditions.

[0048] Through the above processing, the original weighted coupled network is transformed into a sequence of threshold subgraphs with different connection densities. Each threshold subgraph reflects the structural connection relationships preserved under different coupling strengths.

[0049] After obtaining the threshold subgraph, a simplex structure is constructed based on the connectivity between nodes. Specifically, monitoring nodes are considered as zero-dimensional simplexes, any two connected nodes are considered as one-dimensional simplexes, and when multiple nodes form a complete connectivity relationship, a corresponding high-dimensional simplex structure is constructed. This process generates a set of topological structures step by step based on the connection combinations between nodes.

[0050] Subsequently, topological statistical processing is performed on the simple complex structure. First, the number of connected components in the threshold subgraph is counted. The number of connected components describes the number of independent connected regions in the structural response network, denoted as: ;in, Indicates at the threshold The number of connected components under the given conditions. Further, the number of closed connection structures is counted. A closed connection structure consists of multiple nodes forming a closed path; its number characterizes the loop connections formed between nodes, denoted as: ;in, Indicates at the threshold The number of closed connection structures under the given conditions.

[0051] To reflect the overall evolution of structural connection morphology as the threshold changes, the system accumulates the statistical results according to the order of threshold changes. The topological invariant summary feature is obtained by weighted accumulation of topological statistics under different thresholds, and its calculation form is as follows: ; in, This represents the topological invariant summary feature of the p-th class; p represents the topological order. This represents the difference between adjacent thresholds.

[0052] Through the aforementioned multi-threshold evolution processing, the structural information of the dynamic response coupling graph under a single connection state is expanded into a topological statistical description across multiple connection scales, enabling changes in the connection morphology between monitoring nodes to be expressed in a stable topological feature form. The resulting topological invariant summary features are used to characterize the overall change state of the connection relationships in the lightweight steel arch frame structure and provide a unified structural description basis for the subsequent construction of service state features, thereby achieving continuous characterization of the structural response connection morphology.

[0053] Furthermore, in the service status feature construction module, the steps for constructing service status features of lightweight steel arch structures using structural response topology spectrum features and topology invariant summary features include: The topological spectrum features of the structural response are arranged in a unified dimension. The topological invariant summary features are arranged in a unified dimension; The two types of features are combined in a preset order; Generate service status feature vectors for the corresponding time series.

[0054] Specifically, after the topological spectrum decomposition module outputs the structural response topological spectrum feature data, the structural response topological spectrum features are represented as a sequence of feature values ​​sorted by their magnitude. Let's assume that the top M non-zero feature values ​​selected within a certain time window constitute the topological spectrum feature vector, which is represented as: ; in, Indicates at time The structural response topological spectrum eigenvectors; The graph Laplace matrix at time t represents the time t. The m-th eigenvalue; M represents the number of features selected.

[0055] To ensure feature comparability across different time windows, the system arranges topological spectrum features in a unified dimension. Specifically, the same number of feature values ​​are selected in the same order across all time windows, and these values ​​are arranged in ascending order to form a fixed-dimensional feature array. When the number of feature values ​​in a given time window exceeds the preset dimension, only the preset number of feature values ​​are retained; when the number of feature values ​​is insufficient, missing positions are padded with zeros to maintain dimensional consistency.

[0056] After the topological invariant extraction module outputs the topological invariant summary feature data, let's assume at time... The topological invariant summary features are represented as follows: ;in, This represents the cumulative statistical results of connected components; This represents the cumulative statistical results of the closed-loop connection structure. Similarly, the topological invariant summary features are arranged in a uniform dimension. Since the number of topological invariants remains fixed across different time windows, they are directly arranged into a fixed-length vector according to a preset order.

[0057] After unifying the dimensions of the two types of features mentioned above, the service status feature construction module combines them according to a preset order. The combination method is vector concatenation to form a structural service status feature vector: ; in, This indicates that the lightweight steel arch frame is in constant use. , service status feature vector; Represents the structural response topological spectrum eigenvector; This represents the topological invariant summary feature vector.

[0058] After generating the service status feature vector for a single time window, the system arranges the feature vectors at each time point in chronological order to form a service status feature time series: Where K represents the number of time windows within the current analysis period.

[0059] In actual operation, as the sliding time window moves, new structural response data enters the analysis range, while old data is moved out of the time window. The system synchronously updates the topological spectrum features and topological invariant summary features, and reconstructs the corresponding service state feature vector, thereby forming a continuously updated feature time series.

[0060] Through the above steps, graph structure features and topological statistical features are integrated into a fixed-dimensional expression of structural service status, enabling structural correlation states under different time windows to be represented by a consistent data structure. This provides a unified input for subsequent disturbance decoupling and service status parameter calculation, thereby achieving a comprehensive expression of structural response correlation information.

[0061] Furthermore, in the disturbance decoupling module, the steps of performing topological layered projection processing on the service state characteristics of the lightweight steel arch frame structure and generating the intrinsic features of the structure by combining external disturbance data include: Obtain time series data of external disturbances; Establish the correspondence between external disturbance data and service status feature vectors; Construct a perturbation mapping model based on the correspondence; The service state feature vector is decomposed by projection based on the disturbance mapping model; The structural intrinsic features are generated based on the decomposition results.

[0062] The steps for projecting the service state feature vectors based on the perturbation mapping model include: Calculate the directional components of the external disturbance; Calculate the directional components excluding external disturbances; The non-external disturbance direction components are retained as structural intrinsic features.

[0063] Specifically, after the service status feature construction module outputs the service status feature vector of the lightweight steel arch structure, this feature vector simultaneously contains information on the structure's own response changes and response components caused by changes in the external environment. To distinguish between these two types of components, the disturbance decoupling module introduces time series data of external disturbances into the calculation. The external disturbance data is acquired by monitoring devices deployed in the environment surrounding the lightweight steel arch, including ambient temperature data, surrounding rock deformation data, and external load change data. Let's assume that at time... The acquired external disturbance data vector is represented as follows: ; in, Represents the data vector of external disturbances; This indicates that the q-th type of disturbance factor occurs at time t. The measured value; q represents the number of disturbance factors.

[0064] To ensure data consistency, the disturbance decoupling module performs time synchronization processing on the time series of external disturbance data and the time series of service status characteristics, establishing a correspondence between the two types of data under the same time index. Let the lightweight steel arch frame be at time... The service status feature vector is: ;in, This represents the structural service status features formed by the combination of topological spectrum features and topological invariant summary features.

[0065] After establishing the time correspondence, a disturbance mapping model is constructed based on the correlation between external disturbance data and service status feature vectors. This mapping model describes the linear influence of external disturbances on service status characteristics, and its expression is as follows: Where B represents the disturbance mapping matrix; each element of matrix B represents the influence coefficient of different disturbance factors on the service state characteristics of each dimension.

[0066] The disturbance mapping matrix is ​​calculated using corresponding data within a historical time window, enabling external disturbance data to be mapped to the service state feature space. After obtaining the disturbance mapping model, topological layered projection processing is performed on the service state feature vectors. First, the directional components of the external disturbance are calculated, with the following form: ;in, This represents the characteristic component caused by external disturbances. Subsequently, the non-external disturbance direction component is calculated, i.e., the external disturbance component is extracted from the original service state characteristics, and its expression is: ;in, This represents the structural intrinsic features; these features retain only the components generated by changes in the structure's own response. During continuous-time operation, the system performs the aforementioned projection decomposition operation on the service state feature vector at each time step, generating a time series of structural intrinsic features. As the sliding time window updates, the perturbation mapping model is updated synchronously, thereby maintaining the separation between perturbation components and structural intrinsic components.

[0067] Through the above processing, the structural response changes caused by external disturbances are decomposed from the service state characteristics, so that the intrinsic characteristics of the structure can independently reflect the structural state change process of the lightweight steel arch frame itself. This provides a data foundation containing only structural response information for subsequent service state parameter calculations, thereby achieving the separation of structural response from the influence of environmental disturbances.

[0068] Furthermore, in the condition assessment and decision-making module, the steps for generating service condition parameters of the lightweight steel arch frame through the time evolution process of structural intrinsic characteristics include: Constructing time-series trajectories of structural intrinsic features; Calculate the relationship between changes in the intrinsic structural features at adjacent time points; State classification is determined based on the changing relationships; Output the corresponding service status parameters.

[0069] Specifically, after the disturbance decoupling module obtains the intrinsic characteristics of the structure, the system acquires characteristic data that reflects only the response changes of the lightweight steel arch frame structure itself under continuous-time conditions. Let's assume at time... The structural intrinsic characteristics are represented as follows: ;in, This represents the structural intrinsic feature vector at time k, and its dimension is consistent with the service state feature vector. This represents the sampling time corresponding to the k-th time window.

[0070] The state assessment and decision-making module first arranges the structural intrinsic features at each moment in chronological order, forming a time series trajectory of structural intrinsic features: Where K represents the number of time windows within the analysis period.

[0071] After obtaining the time series trajectory, the changing relationships between the structural intrinsic characteristics at adjacent time points are calculated. The changes at adjacent time points are used to characterize the degree of evolution of the structural state under continuous-time conditions, and their calculation form is as follows: ;in, Indicates time The change in structural intrinsic characteristics relative to the previous moment.

[0072] To quantify the overall degree of change, the magnitude of change is calculated from the vector of changes, and its expression is as follows: ;in, This represents the intensity of change in the intrinsic structural features between adjacent time windows. During continuous operation, the system statistically processes the sequence of change in intensity and determines the structural state based on a preset set of state classification thresholds. Let the state classification thresholds be: ;in, This represents the threshold for determining the Lth level state.

[0073] When the intensity of change falls within different threshold ranges, different structural status identifiers are generated. The status assessment and decision-making module determines the service status level of the lightweight steel arch frame at that moment based on a combination of the current intensity of change and the historical time window trend. After determining the status level, the status level is associated with the corresponding time index to form service status parameters, represented as follows: ; in, Indicates time Service status parameters; g() represents the state mapping process based on the intrinsic characteristics and changing relationships of the structure.

[0074] During continuous system operation, as new structural response data enters the analysis process, the time series of intrinsic structural features is updated synchronously. The state assessment and decision-making module continuously calculates new changing relationships and updates service state parameters, thereby achieving continuous assessment of the service state of the lightweight steel arch frame.

[0075] Through the above processing, the evolution process of the intrinsic characteristics of the structure in the time dimension is transformed into quantifiable service status parameters, so that the changes in the structural status can be judged based on continuous time conditions, thereby forming a service status assessment result based on the changes in the structure's own response.

[0076] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. 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 lightweight steel arch service state big data evaluation and decision system, characterized in that, include: The data acquisition module collects structural response data generated by multiple monitoring nodes arranged along the lightweight steel arch frame structure, wherein the monitoring nodes correspond to different structural positions of the lightweight steel arch frame. The data preprocessing module standardizes the structural response data and constructs a sliding time window data sequence in chronological order. The dynamic response coupling diagram construction module calculates the response correlation between monitoring nodes based on the spatial arrangement of monitoring nodes in the lightweight steel arch frame structure and the sliding time window data sequence, and forms a dynamic response coupling diagram characterizing the response correlation state of the lightweight steel arch frame structure. The topology spectrum decomposition module constructs a structural response topology model based on the dynamic response coupling diagram and obtains the corresponding structural response topology spectrum feature data. The topological invariant extraction module obtains topological invariant summary feature data reflecting the connection morphology changes of the lightweight steel arch frame structure based on the multi-threshold evolution process of the dynamic response coupling graph. The service status feature construction module constructs service status features of the lightweight steel arch frame structure based on the structural response topology spectrum features and the topology invariant summary features. The disturbance decoupling module combines external disturbance data to perform topological layered projection processing on the service state characteristics of the lightweight steel arch frame structure and obtains the intrinsic feature data of the structure. The condition assessment and decision-making module obtains the service condition parameters of the lightweight steel arch frame based on the time evolution process of the intrinsic characteristics of the structure.

2. The big data assessment and decision-making system for the service status of lightweight steel arch frames according to claim 1, characterized in that, In the dynamic response coupling graph construction module, the step of forming a dynamic response coupling graph characterizing the response correlation state of the lightweight steel arch frame structure includes: The structural response data of each monitoring node within the sliding time window are synchronized in time. Calculate the time delay correlation between any two monitoring nodes; Calculate the rate-of-change relationship between the two monitoring nodes; The coupling weights between monitoring nodes are determined based on the aforementioned time delay correlation and rate of change synergy. Based on the coupling weights, the connection relationships between monitoring nodes are constructed, and a dynamic response coupling diagram is formed.

3. The big data evaluation and decision system for service state of lightweight steel arch support according to claim 2, characterized in that, The step of constructing the connection relationship between monitoring nodes and generating a dynamic response coupling graph based on the coupling weight includes: Each monitoring node is used as a graph node; Establish weighted connections between nodes according to their coupling weights; The weighted connectivity is updated in time order to form a dynamic response coupling graph sequence.

4. The big data evaluation and decision system for service state of lightweight steel arch support according to claim 1, characterized in that, In the topology spectral decomposition module, the steps of constructing a structural response topology model and generating corresponding structural response topology spectral features from the dynamic response coupling graph include: Establish an adjacency matrix based on the dynamic response coupling graph; Construct the degree matrix based on the adjacency matrix; Generate the graph Laplacian matrix based on the adjacency matrix and degree matrix; Perform eigenvalue decomposition on the graph Laplacian matrix; Based on the decomposition results, the topological spectrum features of the structural response are extracted.

5. The big data evaluation and decision system for service state of lightweight steel arch support according to claim 1, characterized in that, In the topological invariant extraction module, the steps of generating a summary of topological invariant features reflecting the connection morphology changes of the lightweight steel arch frame structure through the multi-threshold evolution process of the dynamic response coupling graph include: The dynamic response coupling diagram is filtered step by step according to multiple preset coupling strength thresholds; Generate corresponding threshold sub-graphs based on each coupling strength threshold; Construct a simple complex structure based on the threshold subgraph; Perform topological statistical processing on the simple complex structure; Generate topological invariant summary features based on statistical results.

6. The big data evaluation and decision system for service state of lightweight steel arch support according to claim 5, characterized in that, The steps for performing topological statistical processing on the simple complex structure include: Count the number of connected components in the threshold subgraph; Count the number of closed connection structures in the threshold subgraph; The statistical results are accumulated according to the order of threshold changes.

7. The big data assessment and decision-making system for the service status of lightweight steel arch frames according to claim 1, characterized in that, In the service status feature construction module, the steps for constructing the service status features of the lightweight steel arch structure using the structural response topology spectrum features and the topology invariant summary features include: The topological spectrum features of the structural response are arranged in a unified dimension. The topological invariant summary features are arranged in a unified dimension; The two types of features are combined in a preset order; Generate service status feature vectors for the corresponding time series.

8. The big data assessment and decision-making system for the service status of lightweight steel arch frames according to claim 1, characterized in that, In the disturbance decoupling module, the step of performing topological layered projection processing on the service state characteristics of the lightweight steel arch frame structure by combining external disturbance data and generating intrinsic structural features includes: Obtain time series data of external disturbances; Establish the correspondence between external disturbance data and service status feature vectors; Construct a perturbation mapping model based on the correspondence; The service state feature vector is decomposed by projection based on the aforementioned disturbance mapping model; The structural intrinsic features are generated based on the decomposition results.

9. The big data assessment and decision-making system for the service status of lightweight steel arch frames according to claim 8, characterized in that, The steps for projecting the service state feature vectors based on the disturbance mapping model include: Calculate the directional components of the external disturbance; Calculate the directional components excluding external disturbances; The non-external disturbance direction components are retained as structural intrinsic features.

10. The big data assessment and decision-making system for the service status of lightweight steel arch frames according to claim 1, characterized in that, In the condition assessment and decision-making module, the steps for generating service condition parameters of the lightweight steel arch frame through the time evolution process of the structural intrinsic characteristics include: Constructing time-series trajectories of structural intrinsic features; Calculate the relationship between changes in the intrinsic structural features at adjacent time points; The state classification is determined based on the aforementioned change relationship; Output the corresponding service status parameters.

Images