An online health monitoring method and system for a steel structure truss

Abstract

The present invention discloses an on-line health monitoring method and system for a steel structure truss, comprising the following steps: Step S1, performing grid processing on the steel structure truss to obtain a set of test grids, taking the center point of each test grid as a test node, and adding a monitoring sensor at each test node; Step S2, sequentially applying a test signal source at each test node, and synchronously acquiring the test monitoring data of the monitoring sensors at all test nodes, and quantifying the influence range of each test node based on the test monitoring data as the criticality of the test node. When measuring, the present invention uses precise calculation to determine the critical nodes of the measurement and set the sensitivity of the monitoring sensors, with strong objectivity, which can improve the accuracy of on-line health monitoring. At the same time, when performing grid processing to determine the critical nodes, an optimization index is used for optimization calculation to obtain the test grid with the optimal index, thereby ensuring the accurate positioning of the critical nodes.

Description

Technical Field

[0001] The present invention relates to the technical field of steel structure truss monitoring, and particularly to an online health monitoring method and system for steel structure trusses. Background Art

[0002] Steel pipe trusses, as a relatively new structural system in roof structures, are applicable to large public building projects. Currently, many large structures in China use this pipe truss structure, and the most widely used is the platform canopy without columns in high-speed railway stations. To meet the requirements of large spaces and unique architectural functions, large-scale, long-span, and complex structural systems such as grid structures, trusses, and cable-supported structures are often adopted.

[0003] In addition, long-span steel structures have relatively small damping and low natural vibration frequencies, and the probability of coupled vibration between the structure and the wind is relatively high. At the same time, due to internal dynamic loads, including vibrations of equipment, vibrations caused by vehicles entering and leaving the station, and vibrations caused by the flow of people and materials in the building, the frequencies of these vibrations are mostly concentrated in the low-frequency region, which is very close to the main frequency of the structure and the main frequency of external wind vibrations. To ensure the structural safety, it is necessary to monitor the vibrations of key parts of the structure and set an upper limit for the vibration amplitude to warn of the structural safety and the normal operation of equipment.

[0004] The prior art CN201910474878.1 discloses an online health monitoring system for long-span pipe truss structures based on optical fiber sensing, which includes an optical fiber grating sensing component, a junction box, a demodulation module, a data processing module, and a monitoring terminal module connected in sequence. The optical fiber grating sensors include optical fiber grating strain sensors, optical fiber grating temperature sensors, optical fiber grating corrosion sensors, optical fiber grating acceleration sensors, and linear detectors, which are arranged on the long-span pipe truss structure to realize the monitoring of multi-state parameters such as strain, vibration, corrosion, deformation, and temperature of the long-span space pipe truss structure, and accumulate monitoring data. The optical fiber grating sensors used in the present invention have the advantages of long measurement distance, anti-electromagnetic interference, easy signal transmission, easy multiplexing, good stability, high measurement accuracy and sensitivity, etc., and have strong practicability and reliability.

[0005] Although the above prior art can ensure the applicability and reliability of online monitoring, it determines the key nodes of measurement and sets the sensitivity of monitoring sensors based on human experience during measurement, with strong human subjectivity, which will lead to a decrease in measurement accuracy and ultimately a decrease in the accuracy of online health monitoring. Summary of the Invention

[0006] The purpose of the present invention is to provide an online health monitoring method and system for steel structure trusses to solve the technical problem in the prior art that when measuring, the key nodes of measurement are determined and the sensitivity of monitoring sensors is set based on human experience, with strong human subjectivity, which will lead to a decrease in measurement accuracy and ultimately a decrease in the accuracy of online health monitoring.

[0007] To solve the above technical problems, the present invention specifically provides the following technical solutions:

[0008] An on-line health monitoring method for a steel structure truss, comprising the following steps:

[0009] Step S1: Perform grid processing on the steel structure truss to obtain a set of test grids, take the center point of each test grid as a test node, and a monitoring sensor is added at each test node;

[0010] Step S2: Apply a test signal source to each test node in turn, and synchronously obtain the test monitoring data of the monitoring sensors at all test nodes, and quantify the influence range of each test node based on the test monitoring data as the criticality of the test node;

[0011] Step S3: Determine key nodes among the test nodes based on the criticality of the test nodes, extract the steel structure attribute parameters of the key nodes, and train a key node recognition model for identifying key nodes in the steel structure truss based on the steel structure attribute parameters and the criticality. The key nodes are characterized as key structures with a wide influence range in the steel structure truss;

[0012] Step S4: Perform grid processing on the steel structure truss to be monitored to obtain a set of test grids and test nodes, obtain the steel structure attribute parameters of each test node, use the key node recognition model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as monitoring nodes, then set monitoring sensors at the monitoring nodes, and set the sensitivity of the monitoring sensors according to the criticality, so as to realize on-line monitoring of dynamic sensitivity at the key structures of the steel structure truss to reduce the redundancy of monitoring data.

[0013] As a preferred solution of the present invention, the grid processing of the steel structure truss to obtain a set of test grids includes:

[0014] Set the basic unit size of the grid, and divide the steel structure truss into a set of basic grids according to the basic grid size;

[0015] Set two mechanical optimization indexes for the steel structure truss. The two optimization indexes include a Young's modulus optimization index and a Poisson's ratio optimization index. The calculation formula of the Young's modulus optimization index is:

[0016]

[0017] The calculation formula of the Poisson's ratio optimization index is:

[0018]

[0019] In the formula, E yThe Young's modulus optimization index characterized as a steel structure truss, μ y The Poisson's ratio optimization index characterized as a steel structure truss, E i The Young's modulus characterized as the i-th basic grid, μ i The Poisson's ratio characterized as the i-th basic grid, n represents the total number of basic grids, and i is a measurement constant;

[0020] Set the thresholds for the Young's modulus optimization index and the Poisson's ratio optimization index. With the maximization of the Young's modulus optimization index and the Poisson's ratio optimization index and being less than the thresholds of the Young's modulus optimization index and the Poisson's ratio optimization index as the constraint conditions, perform expansion and contraction processing on the basic grid size until the basic grid obtained according to the expanded or contracted basic grid size meets the constraint conditions, and then use the basic grid that meets the constraint conditions as the test grid.

[0021] As a preferred embodiment of the present invention, sequentially apply a test signal source at each test node, and synchronously obtain the test monitoring data of the monitoring sensors at all test nodes, including:

[0022] The test signal source sequentially emits test signals starting from each test node, and synchronously obtains the test monitoring data obtained by the monitoring sensors monitoring the test signals at all test nodes;

[0023] Among them, the test signals emitted by the test signal source at different test nodes are random.

[0024] As a preferred embodiment of the present invention, quantize the influence range of each test node as the criticality of the test node based on the test monitoring data, including:

[0025] Compare the similarity between all the test detection data obtained at each test node and the test signal, and compare the similarity between the test detection data and the test signal with the similarity threshold. Among them,

[0026] If the similarity between the test detection data and the test signal is higher than the similarity threshold, then use the test node corresponding to the test detection data as the affected node;

[0027] If the similarity between the test detection data and the test signal is not higher than the similarity threshold, then use the test node corresponding to the test detection data as the non-affected node;

[0028] Count the number of affected nodes of each test node, and use the ratio of the number of affected nodes to the total number of test nodes as the criticality of each test node;

[0029] The calculation formula for the similarity between the test detection data and the test signal is:

[0030]

[0031] wherein, P j,k represents the similarity between the k-th test detection data of the j-th test node and the test signal of the j-th test node, S j,k represents the k-th test detection data of the j-th test node, I j represents the test signal of the j-th test node, and both j and k are measurement constants.

[0032] As a preferred embodiment of the present invention, determining key nodes among the test nodes based on the key degree of the test nodes includes:

[0033] Based on the distribution levels of the test nodes on the z-axis, the test nodes are sequentially divided into multiple layers of test nodes;

[0034] Using MATLAB to perform three-dimensional simulation on the x-y coordinates of the positions of the test nodes and the key degrees of the test nodes in each layer in turn to obtain the three-dimensional key degree model of the test nodes in each layer, and screening out the test nodes corresponding to the key degree peaks in the three-dimensional key degree model as peak nodes;

[0035] Setting a key degree threshold and comparing the key degree of the peak node with the key degree threshold, wherein,

[0036] If the key degree of the peak node is higher than the key degree threshold, the peak node is marked as a key node;

[0037] If the key degree of the peak node is not higher than the key degree threshold, the peak node is not marked as a key node.

[0038] As a preferred embodiment of the present invention, training a key node recognition model for identifying key nodes in a steel structure truss based on the steel structure attribute parameters and key degrees includes:

[0039] Taking the steel structure attribute parameters of the key nodes as the input sequence items of the CNN neural network and the key degrees of the key nodes as the output sequence items of the CNN neural network, and training the CNN neural network based on the input sequence items and output sequence items to obtain the key node recognition model representing the non-linear relationship between the steel structure attribute parameters and key degrees.

[0040] As a preferred embodiment of the present invention, using the key node recognition model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as monitoring nodes includes:

[0041] Sequentially inputting the steel structure attribute parameters of each test node in the steel structure truss to be monitored into the key node recognition model, and outputting the key degree of each test node;

[0042] Compare the criticality with the criticality threshold to screen out the critical nodes of the steel structure truss to be monitored, and use the critical nodes as the monitoring nodes of the steel structure truss to be monitored.

[0043] As a preferred solution of the present invention, setting the sensitivity of the monitoring sensor according to the criticality includes:

[0044] Obtain the criticality of the monitoring node, and set the sensitivity of the monitoring sensor located at the detection node based on the criticality. The setting formula of the sensitivity is:

[0045]

[0046] In the formula, T r represents the sensitivity of the monitoring sensor at the rth monitoring node, and T r,0 represents the basic sensitivity of the detection sensor at the rth monitoring node, and p r represents the criticality of the rth monitoring node.

[0047] As a preferred solution of the present invention, the steel structure truss to be monitored is processed into a grid according to step S1 to obtain a set of test grids and test nodes.

[0048] As a preferred solution of the present invention, the present invention provides a monitoring system according to the on-line health monitoring method of the steel structure truss, including:

[0049] A test module for processing the steel structure truss into a grid to obtain a set of test grids, taking the center point of each test grid as a test node, and adding a monitoring sensor at each test node, applying a test signal source at each test node in turn, and synchronously obtaining the test monitoring data of the monitoring sensors at all test nodes, and quantifying the influence range of each test node based on the test monitoring data as the criticality of the test node;

[0050] A model module for determining critical nodes among the test nodes based on the criticality of the test nodes, extracting the steel structure attribute parameters of the critical nodes, and training a critical node recognition model for identifying critical nodes in the steel structure truss based on the steel structure attribute parameters and the criticality. The critical nodes are characterized as critical structures with a wide influence range in the steel structure truss;

[0051] An application module for processing the steel structure truss to be monitored into a grid to obtain a set of test grids and test nodes, obtaining the steel structure attribute parameters of each test node, using the critical node recognition model to screen the test nodes to obtain the critical nodes of the steel structure truss to be monitored as monitoring nodes, then setting monitoring sensors at the monitoring nodes, and setting the sensitivity of the monitoring sensors according to the criticality.

[0052] The present invention has the following beneficial effects compared with the prior art:

[0053] When measuring, the present invention uses precise calculations to determine the key nodes of the measurement and set the sensitivity of the monitoring sensors. It has strong objectivity, can improve the accuracy of online health monitoring. At the same time, when performing grid processing on the key nodes, an optimization index is used for optimization calculation to obtain the test grid with the optimal index, thereby ensuring the accurate positioning of the key nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, without creative efforts, other implementation drawings can be obtained based on the provided drawings.

[0055] Figure 1 It is a flowchart of the online health monitoring method for a steel structure truss provided by an embodiment of the present invention;

[0056] Figure 2 It is a structural diagram of the monitoring system provided by an embodiment of the present invention.

[0057] The reference numerals in the drawings are respectively represented as follows:

[0058] 1 - Test module; 2 - Model module; 3 - Application module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0060] As Figure 1 shown, the present invention provides an online health monitoring method for a steel structure truss, including the following steps:

[0061] Step S1: Perform grid processing on the steel structure truss to obtain a set of test grids, and use the center point of each test grid as a test node. A monitoring sensor is added at each test node. The monitoring sensors include but are not limited to strain sensors, temperature sensors, vibration sensors, corrosion sensors, etc. The test signal sources include but are not limited to strain force application signal sources, temperature rise and fall application signal sources, vibration application signal sources, corrosion application signal sources, etc. The test monitoring data includes but is not limited to strain force data, temperature data, vibration data, corrosion data, etc.;

[0062] The steel structure truss is meshed to obtain a set of test meshes, including:

[0063] Set the size of the basic unit of the mesh, and divide the steel structure truss into a set of basic meshes according to the basic mesh size;

[0064] Set two mechanical optimization indexes for the steel structure truss. The two optimization indexes include the Young's modulus optimization index and the Poisson's ratio optimization index. The calculation formula of the Young's modulus optimization index is:

[0065]

[0066] The calculation formula of the Poisson's ratio optimization index is:

[0067]

[0068] In the formula, E y represents the Young's modulus optimization index of the steel structure truss, μ y represents the Poisson's ratio optimization index of the steel structure truss, E i represents the Young's modulus of the i-th basic mesh, μ i represents the Poisson's ratio of the i-th basic mesh, n represents the total number of basic meshes, and i is a measurement constant;

[0069] Set the thresholds of the Young's modulus optimization index and the Poisson's ratio optimization index. With the maximization of the Young's modulus optimization index and the Poisson's ratio optimization index and less than the thresholds of the Young's modulus optimization index and the Poisson's ratio optimization index as the constraint conditions, perform expansion and contraction processing on the basic mesh size until the basic meshes obtained according to the expanded and contracted basic mesh size meet the constraint conditions, and then use the basic meshes that meet the constraint conditions as test meshes.

[0070] If the mesh setting is too small, it will cause the mesh unit (test mesh) to not truly reflect the magnitude of the mechanical parameters at that position. If the mesh setting is too large, it will affect the transmission accuracy of the later test signal transmission, and it will also cause the difference between adjacent meshes to become smaller, reducing the test accuracy of key nodes, that is, missing key nodes. Therefore, in this embodiment, using the two optimization indexes of Young's modulus and Poisson's ratio can effectively enable the mesh unit (test mesh) obtained after meshing to solve the above problems, obtain appropriate test meshes, and ensure the accurate positioning of key nodes in the later stage.

[0071] Step S2: Apply test signal sources to each test node in turn, and synchronously obtain the test monitoring data of the monitoring sensors at all test nodes. Quantify the influence range of each test node based on the test monitoring data as the key degree of the test node;

[0072] Apply a test signal source at each test node in sequence and synchronously obtain the test monitoring data of the monitoring sensors at all test nodes, including:

[0073] The test signal source emits test signals starting from each test node in sequence and synchronously obtains the test monitoring data obtained by the monitoring sensors monitoring the test signals at all test nodes;

[0074] Among them, the test signals emitted by the test signal source at different test nodes are random, ensuring that the test signals do not interfere with each other and ensuring the accuracy of the test detection data.

[0075] Quantify the influence range of each test node based on the test monitoring data as the criticality of the test node, including:

[0076] Compare the similarity between all the test detection data obtained at each test node and the test signal, and compare the similarity between the test detection data and the test signal with the similarity threshold. Among them,

[0077] If the similarity between the test detection data and the test signal is higher than the similarity threshold, then the test node corresponding to the test detection data is used as an affected node;

[0078] If the similarity between the test detection data and the test signal is not higher than the similarity threshold, then the test node corresponding to the test detection data is used as a non - affected node;

[0079] Count the number of affected nodes of each test node, and use the ratio of the number of affected nodes to the total number of test nodes as the criticality of each test node;

[0080] The calculation formula for the similarity between the test detection data and the test signal is:

[0081]

[0082] In the formula, P j,k represents the similarity between the k - th test detection data of the j - th test node and the test signal of the j - th test node, S j,k represents the k - th test detection data of the j - th test node, I j represents the test signal of the j - th test node, and j and k are both measurement constants.

[0083] The greater the criticality, the more test nodes can receive the test signals emitted from this test node, indicating that the influence range of this test node is larger. Therefore, the possibility of this test node being a key node is higher.

[0084] Step S3: Determine key nodes among the test nodes based on the key degree of the test nodes, and extract the steel structure attribute parameters of the key nodes. The steel structure attribute parameters include, but are not limited to, the key node angles, stiffness, shear stress, axial compression, etc. Train a key node recognition model for identifying key nodes in the steel structure truss based on the steel structure attribute parameters and the key degree. The key nodes are characterized as key structures with a wide influence range in the steel structure truss.

[0085] Determining key nodes among the test nodes based on the key degree of the test nodes includes:

[0086] Based on the distribution levels of the test nodes on the z-axis, the test nodes are successively divided into multiple layers of test nodes, and the z-axis coordinate values of the test nodes in the same layer are the same;

[0087] Use MATLAB to perform three-dimensional simulations on the x-y coordinates of the positions of the test nodes and the key degrees of the test nodes in each layer in turn to obtain the three-dimensional key degree model of the test nodes in each layer, and screen out the test nodes corresponding to the key degree peaks in the three-dimensional key degree model as peak nodes;

[0088] Set a key degree threshold, and compare the key degree of the peak nodes with the key degree threshold, where,

[0089] If the key degree of the peak node is higher than the key degree threshold, mark the peak node as a key node;

[0090] If the key degree of the peak node is not higher than the key degree threshold, do not mark the peak node as a key node.

[0091] First find the key degree peak, and then perform threshold screening on the key degree peak, which can select the real key nodes among the test nodes with high key degrees adjacent and aggregated, avoid misselection of the test nodes around the key nodes, thereby avoiding the aggregated installation of monitoring sensors, ensuring monitoring safety while saving monitoring costs.

[0092] Training a key node recognition model for identifying key nodes in the steel structure truss based on the steel structure attribute parameters and the key degree includes:

[0093] Take the steel structure attribute parameters of the key nodes as the input sequence items of the CNN neural network, and the key degree of the key nodes as the output sequence items of the CNN neural network. Train the CNN neural network based on the input sequence items and the output sequence items to obtain a key node recognition model that characterizes the non-linear relationship between the steel structure attribute parameters and the key degree.

[0094] Step S4: Mesh the steel structure truss to be monitored to obtain a set of test meshes and test nodes, and acquire the steel structure property parameters of each test node. Use the key node identification model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as the monitoring nodes, then set up monitoring sensors at the monitoring nodes, and set the sensitivity of the monitoring sensors according to the key degree, so as to achieve online monitoring of dynamic sensitivity at the key structures of the steel structure truss to reduce the redundancy of monitoring data.

[0095] Using the key node identification model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as the monitoring nodes includes:

[0096] Input the steel structure property parameters of each test node in the steel structure truss to be monitored into the key node identification model in turn, and output the key degree of each test node;

[0097] Compare the key degree with the key degree threshold to screen out the key nodes of the steel structure truss to be monitored, and use the key nodes as the monitoring nodes of the steel structure truss to be monitored.

[0098] Setting the sensitivity of the monitoring sensor according to the key degree includes:

[0099] Acquire the key degree of the monitoring node, and set the sensitivity of the monitoring sensor located at the detection node based on the key degree. The setting formula of the sensitivity is:

[0100]

[0101] In the formula, T r represents the sensitivity of the monitoring sensor at the r-th monitoring node, T r,0 represents the basic sensitivity of the detection sensor at the r-th monitoring node, and pr represents the key degree of the r-th monitoring node.

[0102] The higher the key degree, the higher the sensitivity is given, so as to achieve more sensitive monitoring of the more critical ones, reduce the time difference between the occurrence of an abnormal situation and its being monitored, and monitor and give an early warning as soon as possible when an abnormality occurs, so as to effectively improve the safety of key nodes and meet the actual monitoring needs.

[0103] The steel structure truss to be monitored is meshed according to step S1 to obtain a set of test meshes and test nodes.

[0104] As Figure 2 shown, based on the above online health monitoring method of the steel structure truss, the present invention provides a monitoring system, including:

[0105] The test module 1 is used to perform grid processing on the steel structure truss to obtain a set of test grids, take the center point of each test grid as a test node, and a monitoring sensor is added at each test node. A test signal source is applied to each test node in sequence, and the test monitoring data of the monitoring sensors at all test nodes is obtained synchronously. The influence range of each test node is quantified based on the test monitoring data as the criticality of the test node;

[0106] The model module 2 is used to determine key nodes among the test nodes based on the criticality of the test nodes, extract the steel structure attribute parameters of the key nodes, and train a key node recognition model for identifying key nodes in the steel structure truss based on the steel structure attribute parameters and the criticality. The key nodes are characterized as key structures with a wide influence range in the steel structure truss;

[0107] The application module 3 is used to perform grid processing on the steel structure truss to be monitored to obtain a set of test grids and test nodes, obtain the steel structure attribute parameters of each test node, use the key node recognition model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as monitoring nodes, then set monitoring sensors at the monitoring nodes, and set the sensitivity of the monitoring sensors according to the criticality.

[0108] When measuring, the present invention uses precise calculations to determine the key nodes to be measured and set the sensitivity of the monitoring sensors, with strong objectivity, which can improve the accuracy of online health monitoring. At the same time, when determining the grid processing of the key nodes, an optimization index is used for optimization calculation to obtain the test grid with the optimal index, thereby ensuring the accurate positioning of the key nodes.

[0109] The above embodiments are only exemplary embodiments of the present application and are not used to limit the present application. The protection scope of the present application is defined by the claims. Those skilled in the art can make various modifications or equivalent replacements within the essence and protection scope of the present application, and such modifications or equivalent replacements should also be regarded as falling within the protection scope of the present application.

Claims

1. An online health monitoring method for a steel structure truss, characterized in that, it includes the following steps: Step S1: Perform grid processing on the steel structure truss to obtain a set of test grids, use the center point of each test grid as a test node, and add monitoring sensors at each test node; Step S2: Apply a test signal source to each test node in sequence, and synchronously obtain the test monitoring data of the monitoring sensors at all test nodes. Quantify the influence range of each test node based on the test monitoring data as the criticality of the test node; Step S3: Determine key nodes among the test nodes based on the criticality of the test nodes, extract the steel structure attribute parameters of the key nodes, and train a key node recognition model for identifying key nodes in the steel structure truss based on the steel structure attribute parameters and the criticality. The key nodes are characterized as key structures with a wide influence range in the steel structure truss; Step S4: Perform grid processing on the steel structure truss to be monitored to obtain a set of test grids and test nodes, obtain the steel structure attribute parameters of each test node, use the key node recognition model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as monitoring nodes, then set monitoring sensors at the monitoring nodes, and set the sensitivity of the monitoring sensors according to the criticality to achieve online monitoring of dynamic sensitivity at the key structures of the steel structure truss to reduce the redundancy of monitoring data; The grid processing of the steel structure truss to obtain a set of test grids includes: Set the size of the basic unit of the grid, and divide the steel structure truss into a set of basic grids according to the basic grid size; Set two mechanical optimization indexes for the steel structure truss, and the two optimization indexes include the Young's modulus optimization index and the Poisson's ratio optimization index; Set the thresholds of the Young's modulus optimization index and the Poisson's ratio optimization index. Taking the maximization of the Young's modulus optimization index and the Poisson's ratio optimization index and being less than the thresholds of the Young's modulus optimization index and the Poisson's ratio optimization index as the constraint conditions, and perform expansion and contraction processing on the basic grid size until the basic grid obtained according to the expanded and contracted basic grid size satisfies the constraint conditions, and then use the basic grid that satisfies the constraint conditions as the test grid.

2. The online health monitoring method for a steel structure truss according to claim 1, characterized in that: The calculation formula of the Young's modulus optimization index is: The calculation formula of the Poisson's ratio optimization index is: where E y represents the optimization index of the Young's modulus of the steel structure truss, μ y represents the optimization index of the Poisson's ratio of the steel structure truss, E i represents the Young's modulus of the i-th basic grid, μ i represents the Poisson's ratio of the i-th basic grid, n represents the total number of basic grids, and i is a measurement constant.

3. The online health monitoring method for a steel structure truss according to claim 1, characterized in that: The step of applying a test signal source to each test node in sequence and synchronously obtaining the test monitoring data of the monitoring sensors at all test nodes includes: The test signal source sequentially emits test signals starting from each test node, and synchronously obtains the test monitoring data obtained by the monitoring sensors monitoring the test signals at all test nodes; Among them, the test signals emitted by the test signal source at different test nodes are random.

4. The online health monitoring method for a steel structure truss according to claim 3, characterized in that: Quantifying the influence range of each test node based on the test monitoring data as the criticality of the test node, including: Comparing the similarity between all the test detection data obtained at each test node and the test signal, and comparing the similarity between the test detection data and the test signal with the similarity threshold, where If the similarity between the test detection data and the test signal is higher than the similarity threshold, the test node corresponding to the test detection data is used as the affected node; If the similarity between the test detection data and the test signal is not higher than the similarity threshold, the test node corresponding to the test detection data is used as the non - affected node; Counting the number of affected nodes of each test node, and taking the ratio of the number of affected nodes to the total number of test nodes as the criticality of each test node; The formula for calculating the similarity between the test detection data and the test signal is: where P j,k represents the similarity between the k-th test detection data of the j-th test node and the test signal of the j-th test node, S j,k represents the k-th test detection data of the j-th test node, I j represents the test signal of the j-th test node, and j and k are both measurement constants.

5. A method for on - line health monitoring of a steel structure truss according to claim 4, characterized in that: Determining the key nodes among the test nodes based on the criticality of the test nodes, including: Based on the distribution level of the test nodes on the z - axis, the test nodes are sequentially divided into multiple layers of test nodes; Using MATLAB to perform three - dimensional simulation on the x - y coordinates of the positions of the test nodes and the criticality of the test nodes in each layer in turn to obtain the three - dimensional criticality model of the test nodes in each layer, and screening out the test nodes corresponding to the criticality peak in the three - dimensional criticality model as the peak nodes; Setting a criticality threshold, and comparing the criticality of the peak nodes with the criticality threshold, where if the criticality of the peak nodes is higher than the criticality threshold, the peak nodes are marked as key nodes; If the criticality of the peak nodes is not higher than the criticality threshold, the peak nodes are not marked as key nodes.

6. A method for on - line health monitoring of a steel structure truss according to claim 5, characterized in that: Training a key node recognition model for identifying key nodes in a steel structure truss based on the steel structure attribute parameters and criticality, including: Taking the steel structure attribute parameters of the key nodes as the input sequence items of the CNN neural network, and the criticality of the key nodes as the output sequence items of the CNN neural network, and training the CNN neural network based on the input sequence items and output sequence items to obtain the key node recognition model representing the non - linear relationship between the steel structure attribute parameters and criticality.

7. A method for on - line health monitoring of a steel structure truss according to claim 6, characterized in that: Using the key node recognition model to screen the test nodes to obtain the key nodes of the steel structure truss to be monitored as the monitoring nodes, including: Sequentially inputting the steel structure attribute parameters of each test node in the steel structure truss to be monitored into the key node recognition model, and outputting the criticality of each test node; Comparing the criticality with the criticality threshold to screen out the key nodes of the steel structure truss to be monitored, and taking the key nodes as the monitoring nodes of the steel structure truss to be monitored.

8. A method for on - line health monitoring of a steel structure truss according to claim 7, characterized in that: Setting the sensitivity of the monitoring sensor according to the criticality, including: Obtain the criticality of the monitoring nodes, and set the sensitivity of the monitoring sensors located at the detection nodes based on the criticality. The formula for setting the sensitivity is as follows: The criticality of r monitoring nodes. where T r represents the sensitivity of the monitoring sensor at the r-th monitoring node, and T r,0 Denoted as the base sensitivity of the detection sensor at the r-th monitoring node, p r Denoted as the 9. A method for online health monitoring of a steel structure truss according to claim 8, characterized in that: The steel structure truss to be monitored is subjected to grid processing according to step S1 to obtain a set of test grids and test nodes.

10. A monitoring system for the online health monitoring method of a steel structure truss according to any one of claims 1-9, characterized in that it includes: A test module for subjecting the steel structure truss to grid processing to obtain a set of test grids, taking the center point of each test grid as a test node, adding monitoring sensors at each test node, sequentially applying a test signal source at each test node, and synchronously obtaining the test monitoring data of the monitoring sensors at all test nodes, and quantifying the influence range of each test node based on the test monitoring data as the criticality of the test node; A model module for determining critical nodes among the test nodes based on the criticality of the test nodes, extracting the steel structure attribute parameters of the critical nodes, and training a critical node recognition model for identifying critical nodes in the steel structure truss based on the steel structure attribute parameters and the criticality. The critical nodes are characterized as critical structures with a wide influence range in the steel structure truss; An application module for subjecting the steel structure truss to be monitored to grid processing to obtain a set of test grids and test nodes, obtaining the steel structure attribute parameters of each test node, using the critical node recognition model to screen the test nodes to obtain the critical nodes of the steel structure truss to be monitored as monitoring nodes, then setting monitoring sensors at the monitoring nodes, and setting the sensitivity of the monitoring sensors according to the criticality. ​

Images