Multi-source transmission tower damage identification method and system based on structural earthquake recurrence

Abstract

The application discloses a kind of multi-source transmission tower damage identification method and system based on structural earthquake recurrence, it is related to power grid disaster prevention and mitigation technical field.The method is first equivalent modeling to transmission tower and applies seismic load;Again, multi-level evaluation unit is divided, and post-earthquake multi-source mechanical response data is collected and normalized;Build evidence reasoning identification model, and the critical value of dangerous grade and attribute weight are adaptively solved by particle swarm optimization algorithm;Fusion multi-source evidence quantization unit dangerous grade, based on Jousselme distance generates continuous danger degree atlas, realizes multi-scale dangerous site positioning and formulates repair priority.The application does not need model training and artificial intervention, and is light, real-time, can accurately identify dangerous site after earthquake, and provides technical support for power grid post-disaster repair and reinforcement.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary technical field of power grid disaster prevention and mitigation, transmission structure safety monitoring and post-disaster intelligent reconstruction, and specifically relates to a method and system for identifying multi-source transmission tower damage based on structural earthquake reproduction. Background Technology

[0002] Transmission towers are the core steel structural support facilities for power grid transmission lines, and are critical infrastructure for maintaining the smooth operation of the energy transmission lifeline and ensuring the smooth functioning of the social economy. Under earthquake disasters, transmission towers undergo an unbalanced process, making them highly susceptible to cascading damage such as member buckling, node failure, and overall instability, directly leading to power grid outages and hindering post-disaster reconstruction. Accurate, rapid, and automatic identification of dangerous parts of transmission towers after an earthquake is a core technological prerequisite for targeted emergency repairs and scientific reinforcement.

[0003] Existing methods for seismic vulnerability analysis and damage mechanics assessment rely heavily on preset damage thresholds, statistical distribution assumptions, and large amounts of sample data. They can only determine the overall structural failure probability and cannot accurately locate specific hazardous areas.

[0004] On the other hand, existing hazardous site identification technologies mostly rely on high-precision finite element modeling and complex mechanical analysis, which require the construction of refined mathematical and physical models and the calibration of a large number of mechanical parameters. This results in high computational costs and cumbersome operation procedures, failing to meet the engineering needs of rapid post-earthquake identification and lightweight support reconstruction.

[0005] Meanwhile, data-driven identification methods that have emerged in recent years rely on large-scale training models and fixed datasets, lack an adaptive generation mechanism for identification rules, are easily affected by subjective parameter settings, and have insufficient robustness and generalization ability, making it difficult to achieve fully automatic identification of dangerous parts without human intervention.

[0006] In summary, existing technologies have certain shortcomings in the lightweight and intelligent identification of dangerous parts of transmission towers after earthquakes. There is a need to propose an identification method that does not require pre-training, can adaptively generate identification rules, and can quickly locate dangerous parts at multiple scales. Summary of the Invention

[0007] The purpose of this technical approach is to overcome the shortcomings of existing technologies or methods, such as complex modeling, reliance on manually preset parameters, and poor generalization, by providing a multi-source transmission tower damage identification method and system based on structural earthquake reconstruction. Relying on lightweight multi-source monitoring data and integrating evidence reasoning and particle swarm optimization algorithms, it can adaptively generate rules and index weights for identifying hazardous parts, enabling automatic identification of multi-scale hazardous parts of the transmission tower, from the overall structure to key components and then to the internal damage zones of those components. This provides technical support for targeted emergency repair and structural reinforcement after power grid imbalance disasters.

[0008] To achieve the above-mentioned technical objectives, the present invention employs the following technical means:

[0009] A multi-source transmission tower damage identification method based on structural earthquake reproduction includes the following steps:

[0010] (1) Perform equivalent modeling of the multi-layer structure of the transmission tower and apply boundary conditions and seismic loads;

[0011] (2) According to the structural design and force transmission path of the transmission tower, the entire tower is divided into multi-level assessment units, and the post-earthquake multi-source mechanical response data of each multi-level assessment unit is collected and normalized.

[0012] (3) Construct a basic model for identifying dangerous parts based on evidence reasoning: Set a set of danger levels, calculate the confidence level of each assessment unit based on each attribute corresponding to each danger level, and assign an initial weight to each attribute;

[0013] (4) With the goal of maximizing the global recognition confidence, an optimization algorithm is used to adaptively solve the critical value of the hazard level and the attribute weight to generate the optimal recognition parameters that are suitable for the current earthquake scenario;

[0014] (5) Substitute the optimal identification parameters from step (4) into the basic model for identifying dangerous parts constructed in step (3), transform the single-attribute risk level confidence into a weighted basic probability mass, integrate multi-source evidence, and quantify the risk level of each multi-level assessment unit.

[0015] (6) Based on the hazard levels of each multi-level assessment unit obtained in step (5), the discrete hazard levels are continuously interpolated using distance metrics to generate a continuous hazard spatial distribution map, thereby locating the multi-scale hazardous parts of the transmission tower and prioritizing repairs according to hazard levels.

[0016] Furthermore, in step (2), the normalization process uses the logarithmic normalization formula.

[0017] Furthermore, in step (3), a piecewise linear function is used to calculate the confidence level of the single-attribute hazard level, and the formula is as follows:

[0018] ;

[0019] In the formula, This is the current attribute number being evaluated; The rating being evaluated; It is the highest level;

[0020] Number the multi-level evaluation units; Used to represent the total number of units in the entire transmission tower being evaluated; and For the first Attributes Level and The critical value for level 3 danger; Representation unit Based on attributes Observed values; Based on attributes The earthquake damage level was assessed as follows: The confidence level.

[0021] Furthermore, in step (3), the attribute weights satisfy the normalization constraint:

[0022] ;

[0023] in, For the first The weight coefficients of each attribute, This represents the total number of attributes.

[0024] Furthermore, in step (4), the critical value of the danger level and the attribute weight are iteratively optimized by the particle swarm optimization algorithm to adaptively generate the optimal identification parameters that are suitable for the unbalanced disaster scenario.

[0025] Furthermore, step (5), which involves fusing multi-source evidence and quantifying the risk level of each multi-level assessment unit, specifically includes the following sub-steps:

[0026] 5A) Substitute the attribute weights obtained from step (4) into the basic model constructed in step (3) to transform the single attribute risk level confidence into a weighted basic probability mass;

[0027] 5B) Employing nonlinear recursive rules of evidence reasoning to fuse multi-attribute evidence;

[0028] 5C) Eliminate accumulated uncertainty and calculate multi-level evaluation units. Final confidence levels for each hazard level:

[0029] ;

[0030] in, for The residual uncertainty resulting from weight dilution after decomposition; For all Individual attribute evaluation level The basic probability quality of fusion; For unit Final confidence level for each hazard level;

[0031] The risk level of the unit is determined based on the maximum confidence criterion.

[0032] Furthermore, in step (6), the continuous hazard level is obtained by interpolation based on the Jousselme distance value. It generates a spatial distribution map of dangerous parts, outputs a list containing component number, location, hazard level, continuous hazard degree, and damage distribution characteristics, and sorts them from high to low hazard degree to determine the emergency repair priority.

[0033] Furthermore, step (1) specifically includes the following sub-steps:

[0034] 1A) Beam elements are used to simulate the bending-torsional coupling stress characteristics of the main and diagonal members;

[0035] 1B) An elastoplastic constitutive model is constructed by combining the Voce nonlinear isotropic hardening model and the Chaboche kinematic hardening model;

[0036] 1C) The connection between the main member and the diagonal member is simulated by MPC hinge method. The bottom of the tower leg is fixed and constrained. Seismic load is applied by dynamic time history analysis method and dynamic equilibrium equation is solved by Newmark-β method.

[0037] This invention further discloses an intelligent identification system for multi-source transmission tower damage based on structural earthquake reproduction, comprising:

[0038] Modeling and load application elements are used to perform multi-layer structural equivalent modeling of transmission towers and apply boundary conditions and seismic loads.

[0039] The evaluation unit division module is used to divide the entire tower into multi-level evaluation units according to the structure and force transmission path of the transmission tower.

[0040] The data acquisition and normalization module is used to acquire post-earthquake multi-source mechanical response data from various multi-level assessment units and perform normalization processing.

[0041] The evidence reasoning model construction module is used to set up a set of hazard levels, calculate the confidence level of each multi-level assessment unit based on each attribute corresponding to each hazard level, and assign an initial weight to each attribute.

[0042] The adaptive optimization solution module is used to adaptively solve the hazard level critical value and attribute weight with the goal of maximizing the global identification confidence, and generate the optimal identification parameters by using an optimization algorithm.

[0043] The evidence fusion and hazard quantification module is used to substitute the optimal identification parameters into the evidence reasoning identification model, convert the single-attribute hazard level confidence into a weighted basic probability mass, and use evidence reasoning rules to fuse multi-source evidence and quantify the hazard level of each multi-level assessment unit.

[0044] The hazardous location and output module is used to perform continuous processing of the hazardous level based on the hazardous level using distance measurement, generate a continuous spatial distribution map of hazardous level, realize multi-scale hazardous location, and formulate emergency repair priority according to the hazardous level.

[0045] Furthermore, the adaptive optimization solution module uses the particle swarm optimization algorithm to iteratively optimize the critical value of the danger level and the attribute weights, and adaptively generates identification rule parameters that are suitable for unbalanced disaster scenarios.

[0046] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0047] First, lightweight and strong real-time performance: This invention does not require a large number of samples for pre-training or manual setting of parameters. It can automatically complete the identification of dangerous parts by relying solely on multi-source lightweight data, which greatly reduces the computing cost and the threshold for engineering applications, and can meet the needs of rapid identification in post-earthquake emergency response.

[0048] Second, adaptive recognition capability: This invention aims to maximize the global recognition confidence. By optimizing the algorithm to adaptively solve the recognition rule parameters, it can automatically adjust the model according to different earthquake scenarios and structural data, overcoming the shortcomings of traditional methods that rely on fixed thresholds or subjective experience.

[0049] Third, multi-scale precise positioning: This invention processes discrete hazard levels into a continuous form and uses distance metric interpolation to generate a spatial distribution map of hazard levels. This enables precise positioning of hazard locations at multiple scales, from the "overall structure" to "key components" and then to "damage zones inside components," providing a more detailed decision-making basis for targeted emergency repairs after disasters.

[0050] Fourth, it has strong engineering versatility: This invention has good versatility and can be extended to the identification of post-earthquake dangerous parts of various power grid steel structures (such as substation frames, communication towers, etc.), and has high engineering promotion value. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the tower head structure of an electric power tower.

[0052] Figure 2 This is a schematic diagram of the tower body and tower legs of an electric power tower.

[0053] Figure 3 An equivalent numerical model of the power tower to be studied;

[0054] Figure 4 This is a schematic diagram of the combined kinematic hardening model for Q235 steel.

[0055] Figure 5 This is a schematic diagram of the combined kinematic hardening model for Q345 steel.

[0056] Figure 6 The east-west acceleration time history of the amplitude-modulated EI Centro wave;

[0057] Figure 7 The north-south acceleration time history of the amplitude-modulated EI Centro wave;

[0058] Figure 8 The vertical acceleration time history of the amplitude-modulated EI Centro wave;

[0059] Figure 9 This is a schematic diagram of the section region division for a beam element;

[0060] Figure 10 A schematic diagram showing the directions x1 and z1 of the shear force and bending moment at the cross section;

[0061] Figure 11 After normalization Distribution cloud map;

[0062] Figure 12 The graph shows the optimization iteration curve of the algorithm.

[0063] Figure 13 A confidence level distribution cloud map for the assessment rating of transmission towers;

[0064] Figure 14 Post-earthquake discrete damage level cloud maps for each participating unit;

[0065] Figure 15 This is a cloud map showing the hazard level of each unit in the continuous transmission tower case study. Detailed Implementation

[0066] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are merely preferred embodiments of the present invention, used to more clearly illustrate the technical solution of the present invention, and do not constitute a limitation on the scope of protection of the present invention. Other implementations made by those skilled in the art based on the concept of the present invention are all within the scope of protection of the present invention.

[0067] The present invention provides an intelligent identification method for multi-source transmission tower damage and hazardous areas based on structural earthquake reproduction, comprising the following steps:

[0068] (1) Equivalent modeling of multi-layered transmission tower structures and application of boundary conditions and seismic loads;

[0069] (2) Division of transmission tower assessment units and normalization of multi-source damage data collection;

[0070] (3) Construct a basic model for identifying dangerous parts based on evidence reasoning;

[0071] (4) The particle swarm optimization algorithm adaptively solves the critical values ​​and attribute weights of the identification rules;

[0072] (5) Multi-attribute evidence fusion and quantitative calculation of hazard measurement of assessment units;

[0073] (6) Identification and location of dangerous parts of transmission towers after earthquake and data support for post-disaster repair.

[0074] This embodiment uses a 220kV double-circuit angle steel transmission tower as an example. The tower is located in an area with a seismic fortification intensity of 8 degrees and a site category of Class II. The main material is Q345 steel, and the diagonal and auxiliary materials are Q235 steel. A refined numerical model is established using ANSYS finite element software. Combined with EI Centro seismic wave triaxial input, an earthquake disaster scenario is simulated, and the method described in this invention is used to achieve intelligent identification of dangerous parts after the earthquake.

[0075] The sub-steps of step (1) in the example are as follows:

[0076] 1A) The tower head structure of the selected double-circuit angle transmission tower, and the tower body and tower leg structure are shown in the attached figure. Figure 1 As shown in Figure 2. Based on the structural symmetry and stress characteristics of the transmission tower, attached... Figure 3 An equivalent numerical model built using beam elements is presented;

[0077] 1B) In the example, the main structural steel used in the transmission tower is Q345 steel, and the auxiliary structural steel is Q235 steel. Both types of steel are constructed using a combination of the Voce nonlinear isotropic hardening model and the Chaboche kinematic hardening model, with the Chaboche kinematic hardening model being a second-order model.

[0078] ;

[0079] ;

[0080] ;

[0081] In the Chaboche kinematic hardening model, for Q235 steel, take... , , , For Q345 steel, take , , , In the Voce nonlinear isotropic hardening model, for Q235 steel, take... , , For Q345 steel, take , , The constitutive models of Q235 steel and Q345 steel using the combined hardening model are shown in the attached figures. Figure 4 and appendix Figure 5 As shown;

[0082] 1C) To address the connection characteristics of transmission tower nodes, the MPC hinge method is used to simulate the connection between the main material and the diagonal material. Fixed constraints are used at the bottom of the tower legs to restrict all degrees of freedom, simulating the actual constraint state between the foundation and the tower feet.

[0083] Note: This case study employs significant simplification in the application of boundary conditions during numerical modeling. Generally, the coupling effects between concrete piles and the surrounding soil under seismic conditions need to be considered, as do the interactions between transmission towers and adjacent towers via conductors. Although these boundary effects are not considered in this case study, their inclusion or exclusion does not affect the subsequent selection of multi-source mechanical response data, the construction of the evidence-based hazard identification model, or the determination of adaptive solution identification rules. This invention emphasizes the innovation of the methodology and framework, rather than the completeness and detail of the numerical modeling.

[0084] 1D) Using dynamic time history analysis, the seismic acceleration time history is analyzed. The loads were converted into equivalent boundary conditions and applied to the model, totaling 820 load steps. (See attached image) Figure 6 Figures 7 and 8 show the EI Centro seismic wave acceleration time histories for a fortification intensity of 8 degrees, after amplitude adjustment and input at a three-dimensional ratio of 1:0.85:0.65.

[0085] The sub-steps of step (2) in the example are as follows:

[0086] 2A) Based on the structural design of the transmission tower, the overall structure of the transmission tower is divided into multiple levels of units;

[0087] 2B) Collect post-earthquake multi-source mechanical response data of the transmission tower. For the transmission tower model constructed with beam elements, the maximum stress in five regions of the beam element section is to be selected: , , , , and z1 turning angle x1 direction turning angle Axial strain Twist angle Bending moment in the z1 direction of the section Bending moment in the x1 direction of the section Shear force in section z1 Shear force in section x1 As multi-source data. (Attached) Figure 9 The beam element section region division is shown, with appendix. Figure 10 This displays the directions x1 and z1 of the cross-sectional shear force and bending moment. It also groups the responses according to their categories:

[0088] Section stress group (combination 1): , , , , ;

[0089] Deformation group (combination 2): , , , ;

[0090] Internal Force Group (Combination 3): , , , Then, select the group that contributes the most to the damage from each group and combine them to form group 4;

[0091] 2C) Collected multi-source raw data Linear normalization is performed, mapping to the [0,1] interval to eliminate differences in units and magnitudes. The normalization formula is:

[0092] ;

[0093] Appendix Figure 11 Showing Normalized cloud map;

[0094] The sub-steps of step (3) in the example are as follows:

[0095] 3A) Define a set of hazard levels The level of danger increases progressively; in this example, a set of danger levels is defined. ;

[0096] 3B) Establish a multi-attribute hazard identification decision model, unit Based on the The confidence level of the risk level of each attribute is expressed as follows:

[0097] ;

[0098] In the formula, ( (Total number of attributes) ( (as indicated by the highest hazard level number) ( (Total number of evaluation units) Based on attributes The earthquake damage level was assessed as follows: Confidence level;

[0099] 3C) The single-attribute hazard confidence level is calculated using a piecewise linear function, and the formula is:

[0100] ;

[0101] In the formula, and For the first Attributes Level and The critical value for level 3 danger;

[0102] 3D) Introducing Attribute Weight Vectors Each attribute represents its contribution to hazard identification, and the weights satisfy the following constraints:

[0103] ;

[0104] This completes the construction of the basic identification model for evidence reasoning. The determination of key parameters of the basic model for identifying dangerous parts in step (3) will be completed in step (4).

[0105] The sub-steps of step (4) in the example are as follows:

[0106] 4A) Construct a particle swarm optimization objective function, with the goal of maximizing global recognition confidence. The formula is:

[0107] ;

[0108] In the formula, For unit Determined as Final confidence level for level 3 danger;

[0109] 4B) When using particle swarm optimization algorithm for automatic key value optimization, the particle position vector Includes the critical values ​​and weights for the danger levels of each attribute, and the particle velocity vector. The iterative update formula is:

[0110] ;

[0111] ;

[0112] For this case, set the total number of iterations. Total number of particles Model coefficients , =1.8, =1.8. After iterative optimization until convergence, the optimal critical values ​​and weights of each attribute are output. The optimization iteration curve of the algorithm is shown in the attached figure. Figure 12 As shown in Table 1, the optimization results for each combination are presented.

[0113]

[0114] The optimization results of this embodiment show that the shear force in the z1 direction of the cross section is... Maximum stress in section 5 z1 turning angle The three categories of indicators that contribute the most to the damage are used to automatically form the optimal identification rules.

[0115] The sub-steps of step (5) in the example are as follows:

[0116] 5A) Convert single-attribute confidence scores into weighted basic probability mass, using the following formula:

[0117] ;

[0118] Unassigned probability quality characterization identifies uncertainty:

[0119] ;

[0120] 5B) Employing nonlinear recursive rules of evidence reasoning to fuse multi-attribute evidence, the first... The formula for fusing individual attributes is:

[0121] ];

[0122] ;

[0123] Normalization factor:

[0124] ;

[0125] 5C) Eliminate accumulated uncertainties to obtain the unit Final confidence levels for each hazard level:

[0126] ;

[0127] Determine the hazard level of the unit based on the maximum confidence criterion:

[0128] ;

[0129] Through the above steps, we can initially obtain the discrete damage assessment results for each participating unit in the transmission tower model. The confidence distribution cloud map of the transmission tower assessment level is attached. Figure 13 As shown in the attached image, the discrete damage level contour maps of each participating unit after the earthquake are as follows. Figure 14 As shown.

[0130] The sub-steps of step (6) in the example are as follows:

[0131] 6A) Select units with a hazard level of 4 to form a candidate set of hazardous locations;

[0132] 6B) The Jousselme distance is used to make discrete hazard levels continuous, thus refining the characterization of differences in hazard levels. The distance formula is:

[0133] ;

[0134] In obtaining each unit Then, interpolation methods can be used to obtain the specific damage values ​​of the element. Figure 15 The map displays a continuous cloud map of hazard levels for each unit. The darkest areas are the connections between the tower legs and the foundation, the connections between the tower legs and the main materials of the lower tower body, and the connections between the main materials inside the tower body. Figure 14 All areas assessed as having a damage level of 4 are the three areas that require the most attention in seismic design or post-earthquake repair.

[0135] 6C) Obtain continuous hazard level based on distance value interpolation Output a list of hazardous parts, including: component number, location, hazard level, continuous hazard level, and damage distribution characteristics.

[0136] Disaster repair priorities are determined by arranging the risks from highest to lowest.

[0137] (i) Extremely dangerous areas: the main materials and connection nodes at the base of the tower legs, which should be replaced and reinforced first;

[0138] (ii) High-risk areas: main tower structure and end members of crossarms, timely reinforcement;

[0139] (iii) Low to medium risk areas: auxiliary materials and secondary components, which can be maintained and treated later;

[0140] The identification results directly provide a basis for decision-making regarding targeted emergency repairs, scientific reinforcement, and rapid recovery of transmission towers after disasters.

[0141] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for identifying multi-source transmission tower damage based on structural earthquake reproduction, characterized in that, Includes the following steps: (1) Perform equivalent modeling of the multi-layer structure of the transmission tower and apply boundary conditions and seismic loads; (2) According to the structural design and force transmission path of the transmission tower, the entire tower is divided into multi-level assessment units, and the post-earthquake multi-source mechanical response data of each multi-level assessment unit are collected and normalized. (3) Construct a basic model for identifying dangerous parts based on evidence reasoning: Set a set of danger levels, calculate the confidence level of each assessment unit based on each attribute corresponding to each danger level, and assign an initial weight to each attribute; (4) With the goal of maximizing the global recognition confidence, an optimization algorithm is used to adaptively solve the critical value of the hazard level and the attribute weight to generate the optimal recognition parameters that are suitable for the current earthquake scenario; (5) Substitute the optimal identification parameters from step (4) into the basic model for identifying dangerous parts constructed in step (3), transform the single-attribute risk level confidence into a weighted basic probability mass, integrate multi-source evidence, and quantify the risk level of each multi-level assessment unit. (6) Based on the hazard levels of each multi-level assessment unit obtained in step (5), the discrete hazard levels are continuously interpolated using distance metrics to generate a continuous hazard spatial distribution map, thereby locating the multi-scale hazardous parts of the transmission tower and prioritizing repairs according to hazard levels.

2. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that, In step (2), the normalization process uses the logarithmic normalization formula.

3. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: In step (3), a piecewise linear function is used to calculate the confidence level of a single attribute hazard level, and the formula is as follows: ； In the formula, This is the current attribute number being evaluated; The rating being evaluated; It is the highest level; Number the multi-level evaluation units; Used to represent the total number of units in the entire transmission tower being evaluated; and For the first Attributes Level and The critical value for level 3 danger; Representation unit Based on attributes Observed values; Based on attributes The earthquake damage level was assessed as follows: The confidence level.

4. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: In step (3), the attribute weights satisfy the normalization constraint: ； in, For the first The weight coefficients of each attribute, This represents the total number of attributes.

5. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: In step (4), the critical value of the danger level and the attribute weight are iteratively optimized by the particle swarm optimization algorithm to adaptively generate the optimal identification parameters that are suitable for the unbalanced disaster scenario.

6. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: Step (5) involves fusing multi-source evidence and quantifying the risk level of each multi-level assessment unit, specifically including the following sub-steps: 5A) Substitute the attribute weights obtained from step (4) into the basic model constructed in step (3) to transform the single attribute risk level confidence into a weighted basic probability mass; 5B) Employing nonlinear recursive rules of evidence reasoning to fuse multi-attribute evidence; 5C) Eliminate accumulated uncertainty and calculate multi-level evaluation units. Final confidence levels for each hazard level: ； in, for The residual uncertainty resulting from weight dilution after decomposition; For all Individual attribute evaluation level The basic probability mass of fusion; For unit Final confidence level for each hazard level; The risk level of the unit is determined based on the maximum confidence criterion.

7. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: In step (6), the continuous hazard level is obtained by interpolation based on the Jousselme distance value. It generates a spatial distribution map of dangerous parts, outputs a list containing component number, location, hazard level, continuous hazard degree, and damage distribution characteristics, and sorts them by hazard degree from high to low to determine the emergency repair priority.

8. The method for identifying multi-source transmission tower damage based on structural earthquake reproduction according to claim 1, characterized in that: Step (1) specifically includes the following sub-steps: 1A) Beam elements are used to simulate the bending-torsional coupling stress characteristics of the main and diagonal members; 1B) An elastoplastic constitutive model is constructed by combining the Voce nonlinear isotropic hardening model and the Chaboche kinematic hardening model; 1C) The connection between the main member and the diagonal member is simulated by MPC hinge method. The bottom of the tower leg is fixed and constrained. Seismic load is applied by dynamic time history analysis method and dynamic equilibrium equation is solved by Newmark-β method.

9. A multi-source intelligent identification system for transmission tower damage based on structural earthquake reproduction, characterized in that, include: Modeling and load application elements are used to perform multi-layer structural equivalent modeling of transmission towers and apply boundary conditions and seismic loads. The evaluation unit division module is used to divide the entire tower into multi-level evaluation units according to the structure and force transmission path of the transmission tower. The data acquisition and normalization module is used to acquire post-earthquake multi-source mechanical response data from various multi-level assessment units and perform normalization processing. The evidence reasoning model construction module is used to set up a set of hazard levels, calculate the confidence level of each multi-level assessment unit based on each attribute corresponding to each hazard level, and assign an initial weight to each attribute. The adaptive optimization solution module is used to adaptively solve the hazard level critical value and attribute weight with the goal of maximizing the global identification confidence, and generate the optimal identification parameters by using an optimization algorithm. The evidence fusion and hazard quantification module is used to substitute the optimal identification parameters into the evidence reasoning identification model, convert the single-attribute hazard level confidence into a weighted basic probability mass, and use evidence reasoning rules to fuse multi-source evidence and quantify the hazard level of each multi-level assessment unit. The hazardous location and output module is used to perform continuous processing of the hazardous level based on the hazardous level using distance measurement, generate a continuous spatial distribution map of hazardous level, realize multi-scale hazardous location, and formulate emergency repair priority according to the hazardous level.

10. The intelligent identification system for multi-source transmission tower damage based on structural earthquake reproduction according to claim 9, characterized in that: The adaptive optimization solution module uses the particle swarm optimization algorithm to iteratively optimize the critical value of the danger level and the attribute weight, and adaptively generate identification rule parameters that are suitable for unbalanced disaster scenarios.

Images