Method and system for evaluating firmness safety of power transmission line iron tower based on computer assistance

Abstract

The invention discloses a firmness safety assessment method and system for a power transmission line iron tower based on computer assistance, and belongs to the technical field of computer-aided engineering, and the method comprises the steps: forming a multi-source data set through fusing sensor, environment and historical maintenance data, and generating preprocessing data after time sequence alignment and filtering processing; and adjusting finite element mechanical parameters based on the data, constructing a dynamic parameterized finite element structure, and performing simulation to obtain physical index data. A preliminary risk index is generated by combining a historical fault library and risk rule matching, then an environment severity degree index is introduced, a comprehensive state index is obtained through calculation, hierarchical feature extraction and multi-scale fusion analysis are carried out, and finally risk early warning information is output. According to the technical scheme, the dynamic parameterization finite element structure is constructed and checked, and comprehensive evaluation and prediction are carried out in combination with data-driven risk matching, so that comprehensive and prospective evaluation and early warning of the firmness safety state of the power transmission line iron tower can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer-aided engineering, in particular to a computer-aided power transmission tower firmness safety evaluation method and system. BACKGROUND

[0002] The power transmission tower is an important part of the power grid system, and its structural firmness is directly related to the safety and stability of power transmission. In order to ensure the safe operation of the tower, its structural health condition needs to be evaluated regularly or in real time. Computer-aided engineering technology, especially the design and simulation method based on finite element analysis, is the core technical means for such structural safety evaluation, which analyzes the stress, strain and stability of the structure by building a digital model of the structure and simulating its mechanical response under various loads.

[0003] In the prior art, the safety evaluation of the power transmission tower mainly relies on field manual inspection and computer-aided simulation analysis. In terms of computer-aided simulation, the usual practice is to establish a geometrically accurate finite element model according to the original design drawings of the tower, and the material parameters of the model adopt the standard ideal values of steel materials, and the boundary conditions are usually simplified as ideal fixed constraints or hinged constraints. In the simulation calculation, the external load applied is usually the standard working condition of extreme wind load, icing load or seismic load specified in the design specification, and through one-time static or dynamic analysis, it is checked whether the bearing capacity of the tower under the design working condition meets the requirements.

[0004] However, the above prior art solution has obvious technical defects in practical application. The finite element model established based on the original design parameters is a static and idealized model, which cannot reflect the structural performance degradation of the tower due to material corrosion and aging, loose connecting bolts, fatigue damage accumulation and uneven foundation settlement, etc. At the same time, the standard design load applied also has significant differences with the random and variable actual environmental load borne by the tower in nature. This state disconnection between the simulation model and the physical entity makes it difficult to accurately reflect the true safety condition of the tower at any time through simulation analysis, which may lead to missed or misjudged risks. SUMMARY

[0005] To solve the above problems, the present application provides a computer-aided power transmission tower firmness safety evaluation method and system, which adopts a technical solution of dynamically building and checking the finite element physical model by fusing multi-source data, and combining data-driven risk matching for comprehensive evaluation and prediction, which can realize accurate, comprehensive and forward-looking evaluation and early warning of the firmness safety state of the power transmission tower.

[0006] The above object can be achieved by the following solution: The computer-aided transmission line tower firmness safety evaluation method and system comprises the following steps: obtaining sensor data, environment data and historical maintenance data of a transmission line tower, fusing and processing the data to obtain a multi-source data set; performing time series alignment and adaptive filtering of abnormal data based on the multi-source data set to generate preprocessed data; adjusting the mechanical parameters of finite element analysis based on the preprocessed data to generate a dynamic parameterized finite element structure; performing simulation calculation based on the dynamic parameterized finite element structure to generate physical index data containing stress distribution and deformation trend; matching the preprocessed data with a preset historical fault library and a preset risk rule to generate a preliminary risk index; determining an environment severity index based on the environment data in the preprocessed data, combining the environment severity index, the physical index data and the preliminary risk index to generate a comprehensive state index; performing hierarchical feature extraction based on the comprehensive state index to separate short-term features and long-term features; performing multi-scale feature fusion and adaptive evaluation analysis on the short-term features and the long-term features to output risk warning information.

[0007] Optionally, the generation of preprocessed data comprises: performing time and space dimension alignment processing on the sensor data, environment data and historical maintenance data in the multi-source data set to generate spatio-temporal alignment data; extracting environment context from the spatio-temporal alignment data and dynamically adjusting the filtering threshold based on the environment context to generate a dynamic filtering threshold; filtering the spatio-temporal alignment data using the dynamic filtering threshold to remove noise and generating preprocessed data.

[0008] Optionally, the generation of a dynamic parameterized finite element structure comprises: comprehensively identifying and establishing a current physical state representation of the tower structure from the preprocessed data; adaptively calibrating the material mechanics parameters in the finite element analysis based on the physical state representation to obtain calibrated material parameters, the material mechanics parameters including elastic modulus, Poisson's ratio, yield strength, fatigue damage accumulation factor and damping characteristics; real-time extracting and quantifying current actual environmental load information from the preprocessed data, dynamically determining the external load application method, loading point and load size according to the actual environmental load information to obtain determined external load; adjusting the boundary conditions and constraint conditions of the finite element analysis in conjunction with the physical state representation and actual environmental load information to obtain adjusted boundary conditions; and generating a dynamic parameterized finite element structure by integrating the calibrated material parameters, the determined external load and the adjusted boundary conditions.

[0009] Optionally, the generating the physical indicator data containing stress distribution and deformation trend comprises: splitting the dynamic parameterized finite element structure into a plurality of calculation units according to physical structure; distributing the calculation units to multi-core processors or cloud computing nodes through a preset task scheduling algorithm, generating distributed calculation tasks; executing the distributed calculation tasks and aggregating calculation results to generate physical indicator data.

[0010] Optionally, the generating the preliminary risk index comprises: extracting current state data from the preprocessed data; matching the current state data with the historical failure library and the risk rules to generate a plurality of risk factors; and performing weighted summation on the plurality of risk factors to generate a preliminary risk index.

[0011] Optionally, the generating the comprehensive state indicator comprises: performing feature extraction based on the environmental data to obtain wind speed, icing thickness, temperature, and humidity, and calculating an environmental severity indicator; dynamically determining weight coefficients of the physical indicator data based on the dynamic parameterized finite element structure through parameter perturbation analysis; calculating a component-level dynamic safety margin based on the calibrated material parameters and the physical indicator data, and combining the dynamic safety margin to normalize and adjust the preliminary risk index to obtain a standardized risk score; weighting the physical indicator data using the weight coefficients, and fusing the standardized risk score and the environmental severity indicator to generate a comprehensive state indicator.

[0012] Optionally, the method further comprises: comparing the preliminary risk index with a simulation risk metric value generated based on the physical indicator data to generate a difference comparison result; generating a conflict signal when the difference comparison result exceeds a preset conflict threshold; and in response to the conflict signal, adjusting parameters of the dynamic parameterized finite element structure according to a preset diagnosis logic library to generate optimized finite element parameters; and performing simulation calculation again using the optimized finite element parameters to generate an updated comprehensive state indicator.

[0013] Optionally, the hierarchical feature extraction based on the comprehensive state indicator comprises: performing sliding window segmentation based on the comprehensive state indicator to extract fluctuation features and trend features within the window as short-term features; and performing historical span analysis based on the comprehensive state indicator to extract degradation features and evolution patterns across time scales as long-term features.

[0014] Optionally, the outputting the risk warning information comprises: fusing the short-term features and the long-term features to form a multi-scale state description; and combining the multi-scale state description and the environmental severity indicator to perform state evolution reasoning to generate a risk level and a warning suggestion in a future period as risk warning information.

[0015] Based on the same inventive concept, the application also provides a computer-aided transmission line tower firmness safety evaluation system, which comprises: a multi-source data acquisition module for acquiring sensor data, environmental data and historical maintenance data of the transmission line tower and performing fusion processing to obtain a multi-source data set; a time series alignment preprocessing module for performing time series alignment and adaptive filtering of abnormal data based on the multi-source data set to generate preprocessed data; a dynamic finite element modeling module for adjusting the mechanical parameters of finite element analysis based on the preprocessed data to generate a dynamic parameterized finite element structure; a simulation calculation and analysis module for performing simulation calculation based on the dynamic parameterized finite element structure to generate physical index data including stress distribution and deformation trend; a risk matching evaluation module for performing corresponding matching with the preprocessed data, a preset historical fault library and a preset risk rule to generate a preliminary risk index; a comprehensive state calculation module for determining an environmental severity index based on the environmental data in the preprocessed data, combining the environmental severity index, the physical index data and the preliminary risk index for analysis to generate a comprehensive state index; a hierarchical feature extraction module for hierarchical feature extraction based on the comprehensive state index to separate short-term features and long-term features; and a multi-scale risk evaluation module for multi-scale feature fusion and adaptive evaluation analysis of the short-term features and the long-term features to output risk warning information.

[0016] Compared with the prior art, the application has the following advantages: 1. The application realizes the accuracy and authenticity of the firmness safety evaluation of the tower by constructing a dynamic parameterized finite element structure. Traditional evaluation methods mostly use static finite element models based on initial design drawings, which have fixed parameters and idealized working conditions, and there is a large deviation from the actual state of the tower in service. The application dynamically calibrates the material mechanics parameters, external load and boundary conditions in finite element analysis by real-time fusion of sensor data, environmental data and historical maintenance data, which can reflect the performance degradation of the tower caused by corrosion, fatigue, foundation settlement and other factors and the real environmental action it bears, improve the fidelity of simulation calculation, and make the physical index data such as stress distribution and deformation trend more accurately represent the real mechanical response of the structure.

[0017] 2, The application fuses the physical simulation and the data-driven two evaluation paths, and constructs a comprehensive and reliable comprehensive evaluation system. Single physical simulation or data analysis method has limitations, the former may be distorted due to model simplification, and the latter lacks deep insight into the structural mechanics mechanism. The application combines high-precision physical index data with the preliminary risk index based on the historical fault library and risk rules, and the physical calibration of the empirical judgment is carried out through the dynamic safety margin, and even the iteration optimization of the model is started when the two results conflict, forming a closed-loop evaluation system that verifies and complements each other. The fusion mechanism makes the evaluation results have the rigor of the physical model and the breadth of the big data analysis, and enhances the reliability and robustness of the evaluation conclusion.

[0018] 3, The application realizes the change from passive monitoring to active early warning by introducing multi-scale feature analysis and state evolution deduction, and significantly improves the forward-looking of safety management. The prior art is mostly limited to the evaluation or alarm of the current state, and lacks the prediction ability of future risks. The application separates the short-term features reflecting sudden risks and the long-term features revealing gradual degradation by hierarchical feature extraction of the comprehensive state index, and combines future environment prediction for state deduction. This method not only can evaluate the instantaneous safety condition of the tower, but also can predict the risk level evolution trend in the future period, and generate early warning suggestions containing specific countermeasures, so that the operation and maintenance management can intervene in advance and prevent problems from happening.

[0019] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0021] Figure 1 is a flowchart of the computer-aided security evaluation method of the power transmission line tower according to an embodiment of the present application.

[0022] Figure 2 is an abnormal data self-adaptive filtering processing diagram according to an embodiment of the present application.

[0023] Figure 3 is a comprehensive state index evolution trend diagram according to an embodiment of the present application.

[0024] Figure 4 is a conflict resolution and model optimization diagram of the embodiment of the present application.

[0025] Figure 5 is a structural schematic diagram of a computer-aided power transmission line tower firmness safety evaluation system according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0027] With reference to Figure 1 , one embodiment of the present application proposes a computer-aided power transmission line tower firmness safety evaluation method, which adopts a technical solution of dynamically constructing and checking a finite element physical model by fusing multi-source data and combining data-driven risk matching for comprehensive evaluation and prediction, and can realize accurate, comprehensive and forward-looking evaluation and early warning of the firmness safety state of the power transmission line tower.

[0028] The method according to the embodiment specifically includes: obtaining sensor data, environmental data and historical maintenance data of the power transmission line tower and performing fusion processing to obtain a multi-source data set; Specifically, the sensor data includes the vibration frequency and the inclination angle of the tower, the environmental data includes the wind speed and the temperature, and the historical maintenance data includes the past maintenance records and the component replacement time. The data fusion adopts a weighted average method, the coefficients of data fusion are determined by expert evaluation according to the reliability and real-time performance of the data source, and the sum of all data fusion coefficients is one, so as to ensure the balance of the fusion result; then the normalized processing is performed on the fused data set to eliminate the dimensional difference, so that all data are in the same numerical range, facilitating subsequent analysis.

[0029] performing time sequence alignment and adaptive filtering processing on the multi-source data set to generate preprocessed data; adjusting the mechanical parameters of the finite element analysis based on the preprocessed data to generate a dynamic parameterized finite element structure; performing simulation calculation based on the dynamic parameterized finite element structure to generate physical index data including stress distribution and deformation trend; corresponding matching is performed between the preprocessed data, the preset historical fault library and the preset risk rules to generate a preliminary risk index; determine an environment severity index based on the environment data in the preprocessed data, analyze the environment severity index, the physical index data, and the preliminary risk index to generate a comprehensive state index; perform hierarchical feature extraction based on the comprehensive state index to separate short-term features and long-term features; perform multi-scale feature fusion and adaptive evaluation analysis on the short-term features and the long-term features to output risk warning information.

[0030] Specifically, first, real-time sensor data, environment data, and historical maintenance records of the power transmission line tower are comprehensively collected, fused, and preprocessed to provide a high-quality unified data basis for analysis. On this basis, two core works are carried out: on the one hand, the preprocessed data are used to dynamically calibrate key parameters such as materials, loads, and boundaries of the finite element model to generate a dynamic parameterized finite element structure that can accurately reflect the current real physical state of the tower, and deep physical index data such as stress distribution and deformation trend inside the structure are obtained through simulation calculation; on the other hand, the preprocessed data are directly used for rapid matching with the built-in historical fault library and risk rules to generate a preliminary risk index based on experience and statistics. Subsequently, the index data obtained by physical simulation, the risk index generated by data-driven, and the quantitative index of the current environment severity are comprehensively calculated to form a global comprehensive state index. Finally, by analyzing the time series of the comprehensive state index, short-term features representing sudden risks and long-term features reflecting gradual degradation are separated, and these multi-scale features are fused and evaluated to finally output forward-looking risk warning information. By constructing a dynamic parameterized finite element structure that can reflect the actual working condition of the tower, the limitations of traditional static ideal models are overcome, and the physical simulation results can more truly reveal the safety margin of the structure. Finally, through multi-scale time domain analysis of the comprehensive state index, not only the current risk can be evaluated, but also the future state evolution trend can be predicted, which changes the safety management from passive fault response to active risk prevention, providing a reliable technical means for ensuring the safe and stable operation of the power transmission line.

[0031] Optionally, the generating preprocessed data comprises: aligning the sensor data, the environment data, and the historical maintenance data in the multi-source data set in time and space dimensions to generate spatiotemporal aligned data; extracting an environment context from the spatiotemporal aligned data and dynamically adjusting a filtering threshold based on the environment context to generate a dynamic filtering threshold; filtering the spatiotemporal aligned data using the dynamic filtering threshold to remove noise to generate preprocessed data.

[0032] Specifically, the first step is to align the multi-source datasets in both time and space dimensions. Time alignment aims to unify the time reference of data from different sources. This involves setting a unified high-frequency sampling time series as the reference. For sensor data or environmental data with sampling frequencies lower than this reference, methods such as linear interpolation or spline interpolation are used for upsampling to generate estimated data at the reference time point. For discrete event-type data such as historical maintenance data, its timestamps are aligned to the nearest reference time point. Spatial alignment aims to associate all data with the specific physical location or components of the transmission line towers. This involves assigning a unique spatial identifier to each sensor, each major pole, or node, mapping the component information involved in the sensor data and historical maintenance data to these unified identifiers. For environmental data, such as regional wind speed or temperature, it is associated with the overall geographical coordinates of the tower. After alignment, spatiotemporally aligned data with consistency is generated. Next, environmental context is extracted based on the spatiotemporally aligned data, and the filtering threshold is dynamically adjusted. Environmental context refers to a set of key environmental parameters that reflect the operating conditions of the tower, such as real-time wind speed, icing thickness, temperature, and humidity extracted from spatiotemporally aligned data. In harsh environments such as strong winds and heavy icing, the tower's normal response signals, such as vibration and strain, are significantly amplified. A fixed filtering threshold may misclassify these valid signals as noise, while in stable environments, it may fail to effectively filter out minute noise. Therefore, a dynamic filtering threshold model is needed. This model correlates the filtering threshold with the environmental context. The calculation of the dynamic filtering threshold... ,have: , in, It is a basic filtering threshold obtained by statistical analysis of historical data under standard or calm environmental conditions, such as taking three times the standard deviation; This represents a vector of environmental context parameters obtained in real time from spatiotemporally aligned data; It is a mapping function used to calculate a dynamic adjustment of the threshold based on the environmental context vector. This function can be obtained through empirical formulas or trained based on a machine learning model, ensuring that the dimensions of its output value are consistent with... The same applies. Finally, the generated dynamic filtering threshold is used to adaptively filter the spatiotemporally aligned data. Each sensor data point in the spatiotemporally aligned data is iterated and compared with the dynamic filtering threshold calculated based on the current environmental context. If the amplitude or rate of change of a data point exceeds the dynamic filtering threshold, it is determined to be abnormal data or noise, and is either removed or repaired by interpolation of adjacent valid data points, thereby obtaining clean and accurate preprocessed data. Figure 2As shown, the original sensor data, i.e. the gray line, contains noise and obvious outliers, such as the spikes in the figure; the dynamic filtering threshold, the shaded interval in the figure, is adaptively adjusted according to the simulated environmental context, and the threshold interval will become wider when the environment is bad. Through the dynamic filtering threshold, abnormal data can be accurately identified and removed / modified, and finally more smooth and accurate pretreatment data, i.e. the solid line part in the figure, is generated. By extracting the environmental context and dynamically adjusting the filtering threshold, adaptive and intelligent filtering of abnormal data is realized, which can accurately distinguish whether the structure is under extreme load or the sensor itself is noisy or faulty according to the actual environmental conditions of the tower, thereby improving the accuracy and reliability of data cleaning and ensuring the authenticity and effectiveness of the data used for subsequent safety evaluation.

[0033] Optionally, the generating the dynamic parameterized finite element structure comprises: comprehensively identifying and establishing a current physical state representation of the tower structure from the pretreatment data; based on the physical state representation, adaptively calibrating material mechanics parameters in the finite element analysis to obtain calibrated material parameters, the material mechanics parameters including elastic modulus, Poisson's ratio, yield strength, fatigue damage accumulation factor and damping characteristics; real-time extracting and quantifying current actual environmental load information from the pretreatment data, dynamically determining external load application mode, loading point and load size according to the actual environmental load information to obtain determined external load; combining the physical state representation and the actual environmental load information, cooperatively adjusting boundary conditions and constraint conditions of the finite element analysis to obtain adjusted boundary conditions; comprehensively generating the dynamic parameterized finite element structure based on the calibrated material parameters, the determined external load and the adjusted boundary conditions.

[0034] Specifically, first, the current physical state representation of the tower structure needs to be comprehensively identified and established based on the preprocessed data. The physical state representation is a comprehensive description of the actual health status of the tower, which is constructed by analyzing the sensor data and historical maintenance data in the preprocessed data. For example, the stress concentration area and potential plastic deformation of the component are identified using strain sensor data; the inherent frequency and mode shape changes of the tower are analyzed using vibration sensor data to infer the degradation of structural stiffness; whether the foundation has uneven settlement is determined using tilt sensor data; at the same time, combined with the records of component replacement, corrosion coating repair or fastener reinforcement in the historical maintenance data, a complete physical state image reflecting material aging, structural damage and maintenance history is formed. Subsequently, based on the physical state representation, the material mechanics parameters in the finite element analysis are adaptively calibrated to obtain the calibrated material parameters. For the elastic modulus, it can be adjusted according to the actual measured stress-strain relationship to reflect the stiffness reduction of the material due to corrosion or fatigue; the fatigue damage accumulation factor can be extracted from the strain sensor data by the rain flow counting method to obtain the stress cycle number and amplitude, and the Palmgren-Miner linear cumulative damage criterion is used for calculation. For the calculation of the fatigue damage accumulation factor there is: , wherein, is the actual stress cycle number obtained from the preprocessed data at the stress level of the first stress level; is the allowable stress cycle number of the material at this stress level, which is obtained according to the standard stress-life curve of the material. The calculated The values can be used to correct the yield strength and other parameters of the material. For damping characteristics, the actual damping ratio of the current structure can be estimated by analyzing the free vibration decay signal of the tower under wind load excitation and using identification techniques such as the logarithmic decrement method. Then, the current actual environmental load information is extracted and quantified in real time from the preprocessed data to dynamically determine the external load. Specifically, the real-time wind speed and wind direction in the environmental data are converted into wind pressure acting on each member of the tower according to the relevant load specifications; the ice thickness data are converted into the mass and volume added to the members and ground wires, so as to calculate the increased gravity load and wind load windward area. These calculated loads are accurately applied to the corresponding loading points of the finite element structure, and the application manner and load size are determined in the form of distributed or concentrated forces, so as to obtain the determined external load. At the same time, the boundary conditions and constraint conditions of the finite element analysis are adjusted coordinately. Based on the information about foundation settlement or rotation in the physical state representation, the ideal fixed constraint at the bottom of the model is modified to a forced displacement constraint with a specific displacement or a spring element is introduced to simulate the flexibility of the foundation, i.e. the interaction between soil and structure. Combined with the actual environmental load information, if the external load is huge, the nonlinear boundary effect may need to be considered, such as the local voiding that may occur under the action of uplift force of the simulation tension foundation, and the adjusted boundary conditions are obtained through this step. Finally, the calibrated material parameters, the determined external load and the adjusted boundary conditions are integrated into the initial finite element geometric model of the tower, so as to generate a dynamic parameterized finite element structure that can reflect the current material performance of the tower, bear the actual environmental action and consider the real constraint state. This dynamic parameterized modeling method improves the accuracy and reality of the finite element simulation analysis, so that the subsequent stress distribution and deformation trend calculation results can more truly predict the mechanical behavior of the tower.

[0035] Optionally, the generating physical index data containing stress distribution and deformation trend comprises: splitting the dynamic parameterized finite element structure into multiple calculation units according to the physical structure; allocating the calculation units to multi-core processors or cloud computing nodes through a preset task scheduling algorithm to generate allocated calculation tasks; executing the allocated calculation tasks and aggregating the calculation results to generate physical index data.

[0036] Specifically, first, the complete dynamic parameterized finite element structure is decomposed, i.e. according to the physical structure such as tower legs, tower body, cross arm and foundation, etc. main parts, it is divided into a plurality of relatively independent sub-domains in calculation, which are calculation units. This decomposition aims to transform a huge and complex overall calculation problem into a series of smaller sub-problems that can be processed in parallel, laying the foundation for subsequent distributed computing. Then, through a preset task scheduling algorithm, these calculation units are efficiently allocated to available computing resources, i.e. multi-core processors of local computers or remote cloud computing nodes. The core goal of the task scheduling algorithm is to achieve load balancing, ensuring that the computing load of each processor or node is roughly equivalent, thereby minimizing the waiting time and total time consumption of the overall simulation calculation. For example, a scheduling strategy based on calculation cost estimation can be used, in which the cost of each calculation unit is related to the number of units it contains, the number of nodes and the degree of nonlinearity. The scheduler dynamically allocates calculation units according to this cost estimation and the current load of each computing node, generating a series of allocated calculation tasks, each corresponding to the solution process of a calculation unit on a specific node. Finally, each processor or cloud computing node independently and in parallel executes the calculation task assigned to it, i.e. performs finite element solution on the calculation units it is responsible for, calculating the stress, strain and displacement within each calculation unit. After all parallel calculation tasks are completed, the calculation results of each calculation unit are aggregated. The aggregation process includes integrating the stress and strain fields of all units and splicing them into a complete tower deformation pattern according to the displacement coordination conditions and force balance conditions at the boundaries of the sub-domains. Through this step, the final physical index data that can completely describe the mechanical response of the entire tower under the current dynamic parameters and loads is obtained. The physical index data specifically represents the stress distribution cloud map of each component of the tower and the overall deformation trend such as displacement and rotation angle. By introducing parallel computing mechanism, the calculation efficiency of complex transmission line tower finite element simulation analysis is improved, the time required from model establishment to simulation result is shortened, and high-precision dynamic mechanical analysis of the tower becomes a near real-time operation.

[0037] Optionally, the generating the preliminary risk index comprises: extracting current state data from the preprocessed data; matching the current state data with the historical failure library and the risk rules to generate a plurality of risk factors; performing weighted summation on the plurality of risk factors to generate a preliminary risk index.

[0038] Specifically, first, the current state data reflecting the real-time state of the power transmission tower is extracted from the preprocessed data. The current state data is a slice of the preprocessed data at the current time point or a very short time window, including real-time or recent statistical values obtained from various sensors, such as the average and peak strain of key members, the vibration amplitude and main frequency of the tower top, and the inclination angle of the tower foundation, etc.; at the same time, it also includes the current environmental parameters, such as wind speed and temperature; and it also includes relevant historical maintenance data information, such as the time since the last comprehensive maintenance. After obtaining the current state data, it is matched with the pre-set historical failure library and risk rules to generate multiple risk factors. The historical failure library is a structured knowledge base that stores a large number of real tower failure or failure cases that have occurred in history, and each case is associated with the state data characteristics before the occurrence. The matching process is to compare the current state data with the failure precursor characteristics in the historical failure library, and if the similarity exceeds the failure threshold, a risk factor corresponding to the historical failure will be triggered; the failure threshold is a critical value determined by recording the state data such as strain, vibration frequency, and inclination angle before the tower failure in history and statistically analyzing the distribution characteristics of these data before the failure. The risk rule is a series of logical judgment conditions formulated by domain experts or obtained through data mining, such as "if the corrosion depth of a certain main material component exceeds 80% of the design allowable value, a high-level material failure risk factor is generated". By logically judging the current state data one by one with these rules, a series of corresponding risk factors can also be generated. Each risk factor is a quantitative value representing the danger level in a specific risk dimension. Finally, the multiple risk factors generated by the above process are weighted and summed to calculate the comprehensive preliminary risk index. For the calculation of the preliminary risk index , there are: , wherein, represents the risk factor, which is generated by the matching and judgment process described above, and is usually normalized to a standard interval, and the value is obtained by matching similarity or logical judgment; is the weight coefficient corresponding to the risk factor, which reflects the importance of different risk factors to the overall security of the tower, and the value is determined by expert experience, analytic hierarchy process or statistical analysis based on historical data, and the sum of all weight coefficients is one. Through this weighted summation operation, risk information from different dimensions and different properties is integrated into a single and intuitive preliminary risk index. The potential risks of the tower can be quickly identified and quantified, providing a quick and empirical basis for the overall security of the tower.

[0039] Optionally, the generating the comprehensive state indicator comprises: Based on the environmental data, feature extraction is performed to obtain wind speed, icing thickness, temperature, and humidity, and an environmental severity index is calculated; Based on the dynamic parameterized finite element structure, the weight coefficients of the physical indicator data are dynamically determined through parameter perturbation analysis; Based on the calibrated material parameters and the physical indicator data, a dynamic safety margin at the component level is calculated, and the preliminary risk index is normalized and adjusted in combination with the dynamic safety margin to obtain a standardized risk score; The physical indicator data is weighted using the weight coefficients, and is fused with the standardized risk score and the environmental severity index to generate a comprehensive state indicator.

[0040] Specifically, first, the key feature quantities, i.e., wind speed, icing thickness, temperature, and humidity, are extracted from the environmental data. Then, an environmental severity index is calculated using a multi-factor evaluation model. For calculating the environmental severity index using the multi-factor evaluation model ,

[0041] wherein, wind speed, icing thickness, temperature, and humidity obtained from the environmental data, respectively; , , , is a set of normalization functions that map physical quantities with different dimensions to a unified, dimensionless severity score interval, for example, actual wind speed is converted to wind pressure rating score according to relevant design specifications or historical statistical data; , , , are corresponding weight coefficients, reflecting the relative influence degree of different environmental factors on the tower safety, whose values can be determined by expert scoring method or analytic hierarchy process. Then, based on the established dynamic parameterized finite element structure, the weight coefficients of the physical index data are dynamically determined through parameter perturbation analysis. Parameter perturbation analysis is a sensitivity analysis technique, the core of which is to evaluate the response degree of model output to the small change of input parameters. Here, the key input parameters of the finite element model, such as the size of external load or the elastic modulus of material, are slightly perturbed, and then simulation calculation is performed to observe the change amplitude of the physical index data, such as the stress of a specific component or the displacement of a key node. For those physical indexes with more dramatic response changes, higher weight coefficients are given because they are the key and sensitive indicators reflecting the change of structure state. Subsequently, the risk index is physically checked and standardized. First, based on the physical index data obtained from simulation calculation and the calibrated material parameters already included in the finite element model, the dynamic safety margin at the component level is calculated. The dynamic safety margin is defined as the ratio of the bearing capacity of the component to the actual load effect it bears. Then, the preliminary risk index generated previously is normalized and adjusted to generate a standardized risk score using the calculated minimum dynamic safety margin of the whole tower. This adjustment process aims to correct the preliminary judgment based on experience and rules with the real results of physical simulation, such as when the minimum dynamic safety margin is extremely low, even if the preliminary risk index is not high, it should be adjusted upward, and vice versa. Finally, all information is fused to generate a comprehensive state indicator. For calculating the comprehensive state indicator , , wherein, is the weight coefficient of the i-th physical index obtained by parameter perturbation analysis, which is determined by slightly perturbing the key input parameters such as load and material parameters, and observing the change amplitude of the output physical index such as stress and displacement; is the dimensionless score of the original physical index data such as stress value, for example, obtained by dividing the actual stress by the allowable stress; represents the structure health score based on physical simulation. is the standardized risk score after dynamic safety margin check. , and are the fusion weights of the two scores, whose sum is one, used to balance the proportion of physical simulation results, standardized risk score and environment severity index in the final indicator. For example, Figure 3 ​As shown, different risk areas of "normal", "attention" and "risk" are marked in the figure, and are distinguished by color from light to dark; the up and down fluctuations and overall trend of the comprehensive state index curve reflect the dynamic evolution of the tower safety state, once the comprehensive state index crosses the preset attention threshold or risk threshold, the corresponding alarm can be sent out to realize situation awareness. Through this calculation, a comprehensive state index is finally obtained which takes into account the physical reality and experience knowledge. The generated comprehensive state index has both the rigor and depth of the physical model, and also absorbs the breadth of historical data and expert knowledge, so that the final evaluation result is more comprehensive, reliable and accurate compared to a single method.

[0042] Optionally, the method further comprises: differentially comparing the preliminary risk index with the simulated risk metric value generated according to the physical index data, and generating a conflict signal when the differential comparison result exceeds a preset conflict threshold; in response to the conflict signal, adjusting the parameters of the dynamic parameterized finite element structure according to a preset diagnostic logic library to generate optimized finite element parameters; re-performing simulation calculation using the optimized finite element parameters to generate an updated comprehensive state index.

[0043] Specifically, first, the risk assessment results of two different sources need to be compared for differences. One is the preliminary risk index, which is generated based on preprocessed data, historical failure library and risk rules, representing the judgment results of data-driven and experience knowledge. The other is the simulation risk measurement value generated according to the physical index data. This simulation risk measurement value is a kind of risk interpretation of the simulation results, which can be defined as the ratio of the maximum stress of the whole tower to the material yield strength, or the reciprocal of the minimum dynamic safety margin. After normalizing the two risk assessment results, the absolute difference or relative difference between them is calculated. If the difference comparison result exceeds a preset conflict threshold, a conflict signal will be generated; the conflict threshold is usually determined by experts according to the system sensitivity requirements and historical false alarm rate. For example, when the preliminary risk index shows low risk, but the simulation risk measurement value shows high risk, it is considered that there is a significant conflict, and a conflict signal will be generated and responded to, starting the automatic adjustment process of the finite element parameters. The diagnostic logic library is a knowledge base containing expert knowledge and reasoning rules, which associates different types of conflict patterns with possible problems of finite element parameters. For example, if the simulation stress is much higher than the level reflected by the sensor measured strain, the diagnostic logic library may infer that the external load parameter in the model, such as the wind load coefficient, is set too high, or the elastic modulus in the material mechanics parameter is not accurate. According to the guidance of the diagnostic logic library, one or more parameters in the dynamic parameterized finite element structure will be automatically adjusted, such as reducing the wind load coefficient or correcting the elastic modulus, to generate a set of optimized finite element parameters. Finally, using this set of optimized finite element parameters, a complete simulation calculation process is re-initiated. This means that based on the adjusted parameters, the simulation calculation is performed again to generate new physical index data, and the dynamic safety margin is recalculated based on the new physical index data, and then the standardized risk score is updated to finally generate an updated comprehensive state indicator. For example Figure 4 As shown in FIG. 8, the difference comparison result is represented by the absolute difference in the figure; the “preliminary risk index” is relatively stable, while the “simulation risk measurement value” has a large difference with the preliminary risk at the initial time, exceeding the “conflict threshold”; after detecting the conflict, model parameter optimization is performed, so that the simulation risk measurement value gradually converges in the subsequent iterations, and tends to be consistent with the preliminary risk index. This process is an iterative optimization cycle, until the difference between the preliminary risk index and the simulation risk measurement value converges within the conflict threshold, or the maximum number of iterations is reached. This closed-loop feedback correction process ensures that the fidelity of the dynamic parameterized finite element structure can be continuously improved during use, so that the model can more accurately reflect the true state of the tower, and ensure the continuous effectiveness and accuracy of the safety evaluation results.

[0044] Optionally, the hierarchical feature extraction based on the comprehensive state indicator includes: performing sliding window segmentation based on the comprehensive state indicator, extracting fluctuation features and trend features in the window as short-term features; performing history span analysis based on the comprehensive state indicator, extracting degradation features and evolution patterns across time scales as long-term features.

[0045] Specifically, first, in order to capture the short-term dynamic changes of the tower state, sliding window segmentation is performed based on the time series of the comprehensive state indicator. That is, a fixed length time window is set, such as several hours or several days, and the window is slid along the time series of the comprehensive state indicator. For each data segment in the window, statistical features are calculated to represent the short-term fluctuations and trends. Fluctuation features can be quantified by calculating the standard deviation, variance or coefficient of variation of the data in the window, reflecting the instability of the tower state in a short time or the degree of influence by sudden events. Trend features can be obtained by linear regression analysis of the data in the window, extracting the slope, or calculating the change rate of the values at the start and end of the window, which represents the monotonic change direction of the state indicator in the short term, such as continuous deterioration or improvement. These fluctuation features and trend features extracted from the sliding window together constitute the short-term features of the tower state. Secondly, in order to reveal the long-term performance evolution of the tower structure, history span analysis is performed based on the comprehensive state indicator. This analysis focuses on a longer time scale, such as several months, several years or even the entire service period. That is, the comprehensive state indicator sequence accumulated over a long period of time is analyzed to extract degradation features and evolution patterns across time scales. Degradation features can be obtained by fitting a high-order polynomial to the entire historical sequence or applying a nonlinear regression model. The parameters of the nonlinear regression model can quantify the overall decay rate and pattern of the tower performance over time, such as identifying whether the structure is in an accelerated aging phase. The extraction of evolution patterns can be achieved by time series decomposition techniques, such as seasonal decomposition, which decomposes the indicator sequence into trend, seasonal and residual terms. The trend term reflects the long-term evolution pattern, while the seasonal term may reveal periodic risk patterns related to annual climate changes. These degradation features and evolution patterns obtained through analysis constitute the long-term features of the tower state. This hierarchical extraction method enables the subsequent risk assessment to not only respond to sudden conditions, but also to conduct forward-looking health management.

[0046] Optionally, the output risk warning information includes: fusing the short-term features and the long-term features to form a multi-scale state description; combining the multi-scale state description and the environment severity indicator to perform state evolution deduction, generating a risk level and a warning suggestion in a future period as risk warning information.

[0047] Specifically, first, the previously isolated short-term features and long-term features need to be effectively fused. This fusion process aims to construct a multi-scale state description that can reflect both the current instantaneous risk and the future evolution trend of the tower. The fusion can splice the short-term features and the long-term features into a high-dimensional feature vector, or use neural network structures such as attention mechanisms to dynamically learn and assign different weights to features of different scales, and then perform weighted fusion, thereby generating a multi-scale state description that can comprehensively depict the current and historical health status of the tower. Then, based on this multi-scale state description and in combination with the currently calculated environment severity index, state evolution deduction is performed. This step aims to predict the safety state change trend of the tower in the future period of time. The deduction can use a data-driven prediction model, such as a long short-term memory network (LSTM), which can learn long-term dependencies in time series data. The current multi-scale state description and the environment severity index are taken as inputs to directly predict the comprehensive state index in the future period of time and generate risk warning information containing the risk level and warning suggestions in the future period of time. According to the preset risk threshold, the predicted comprehensive state index is classified. For example, the comprehensive state index value is divided into normal, attention, risk, and different levels. Once the risk level at a future time point is predicted to exceed the "attention" level, the warning mechanism is triggered. The generation of the warning suggestion is more in-depth, as it not only indicates the risk level but also analyzes the main factors leading to the risk. If the risk is mainly caused by the combination of severe fluctuations in short-term features and adverse environment, the warning suggestion may be "strong wind in the next 24 hours, please pay attention to monitoring the tower vibration and prepare for emergencies"; if the risk is caused by the continuous deterioration of long-term features, the suggestion may be "structural fatigue damage has reached the critical value, it is recommended to arrange detailed inspection and reinforcement within a month". These specific risk levels and warning suggestions together constitute the final output of the risk warning information. By fusing multi-scale features and performing state evolution deduction, the safety assessment is elevated from "post-diagnosis" and "real-time monitoring" to "pre-prediction", realizing the transition from passive response to active defense, providing clear and operable guidance for operation and inspection personnel, and enabling them to take preventive measures in advance to effectively reduce potential failures and accidents.

[0048] Based on the same inventive concept, as shown in Figure 5 The present application also provides a computer-aided security assessment system for the stability of a power transmission tower, which comprises: A multi-source data acquisition module for acquiring sensor data, environmental data, and historical maintenance data of the power transmission tower and performing fusion processing to obtain a multi-source data set; A time series alignment preprocessing module for time series alignment and adaptive filtering of abnormal data based on the multi-source data set to generate preprocessed data; a dynamic finite element modeling module, configured to adjust mechanical parameters of finite element analysis based on the preprocessed data, and generate a dynamic parameterized finite element structure; a simulation calculation analysis module, configured to perform simulation calculation based on the dynamic parameterized finite element structure, and generate physical index data containing stress distribution and deformation trend; a risk matching evaluation module, configured to perform corresponding matching with the preprocessed data, a preset historical failure library and a preset risk rule, and generate a preliminary risk index; a comprehensive state calculation module, configured to determine an environment harshness index based on environment data in the preprocessed data, and perform analysis on the environment harshness index, the physical index data and the preliminary risk index, and generate a comprehensive state index; a hierarchical feature extraction module, configured to perform hierarchical feature extraction based on the comprehensive state index, and separate short-term features and long-term features; a multi-scale risk evaluation module, configured to perform multi-scale feature fusion and adaptive evaluation analysis on the short-term features and the long-term features, and output risk warning information.

[0049] In order to verify the feasibility of the application in implementation, the application is applied to a 500kV super-high voltage transmission line of a certain company. Part of the line section passes through mountainous areas, and is always faced with adverse weather conditions such as strong wind, icing and sudden temperature change. The firmness and safety of the tower structure face severe challenges, and the traditional manual inspection and regular detection method is difficult to realize timely and accurate risk assessment.

[0050] In order to verify the effectiveness of the application, a typical straight tower is selected as a test object on the line, and the evaluation system of the application is deployed. The system collects operation data for one year, covers complete seasonal climate change, and performs detailed firmness and safety evaluation, risk warning and model self-optimization test, and records data under various extreme working conditions such as typhoon, cold wave and freezing during the period.

[0051] In the embodiment, the application first obtains sensor data, environmental data and historical maintenance data by deploying strain gauges, accelerometers and tilt sensors at locations such as tower bases, tower legs and key crossarms, combined with micro-meteorological stations along the line and historical maintenance databases. For example, during a typhoon passage in August 2024, vibration data with a sampling frequency of 50Hz and wind speed data with a frequency of 1Hz were aligned on a unified time reference through an interpolation algorithm. At the same time, according to the environmental context, the instantaneous wind speed at that time reached 35m / s, and the system adaptively increased the filtering threshold of strain data from the conventional 3 times standard deviation to 5 times standard deviation, thereby accurately removing the random noise of the sensor signal and retaining the true response data of the structure under strong wind excitation, generating high-quality preprocessed data.

[0052] Subsequently, the system dynamically constructs a finite element structure based on the preprocessed data. After analyzing the strain data for the entire year of 2024, the system uses the rainflow counting method and the Palmgren-Miner linear cumulative damage criterion to calculate that the fatigue damage accumulation factor of a main material connection node has reached 0.28, and accordingly reduces the yield strength parameter of the material near the node. At the same time, combined with the 0.05 degree uneven settlement of the tower base monitored by the tilt sensor, the ideal fixed constraint at the bottom of the model is modified to an elastic constraint with forced displacement. The simulation calculation task of this dynamically parameterized finite element structure is decomposed and distributed to multiple core computing nodes in the cloud through a task scheduling algorithm, and the stress distribution and deformation trend analysis of the entire tower is completed within 10 minutes, generating physical index data.

[0053] In the risk assessment process, the system matches real-time state data with the historical failure library. For example, on November 28, 2024, the system monitored that the strain value of a component abnormally increased in a low temperature environment, which was highly similar to the "early signs of low temperature brittle fracture" recorded in the historical failure library, and immediately generated a high-level material risk factor. By weighting and summing multiple risk factors, a preliminary risk index of 0.65 was generated. Then, based on the minimum dynamic safety margin of 2.1 of the entire tower obtained through simulation, the system adjusts and checks the preliminary risk index, and combines the environmental severity index and physical index data to finally generate a comprehensive state index of 0.72.

[0054] The closed-loop feedback mechanism of the present application also exhibits its superiority. In a simulation test, the preliminary risk index generated by the system is 0.4, i.e. low risk, while the simulation risk measurement value based on the ratio of maximum stress to yield strength is as high as 0.8, i.e. high risk, with a difference of more than 0.2, the conflict threshold. The system then triggers a conflict signal and makes an inference based on the diagnostic logic library that the conflict reason may be that the icing uneven distribution coefficient in the icing load model is too conservative. The system automatically adjusts the parameter and re-simulates, and the updated comprehensive state index generated has an improved consistency of 30% with the subsequent measured data, realizing self-optimization of the evaluation model.

[0055] Through hierarchical feature extraction, the present application successfully separates short-term and long-term features. Short-term feature analysis reveals a strong correlation between tower vibration amplitude and daily afternoon valley wind. Long-term feature analysis identifies a slight acceleration trend in the settlement rate of the southeast side of the tower foundation by fitting the trend of the annual comprehensive state index. Combined with the weather forecast of heavy rain in the next week, the system uses state evolution to predict that the tower foundation settlement risk will reach the "risk" level after 72 hours, and generates the warning information "suggestion to focus on the soil moisture content of the southeast side of the tower foundation and make temporary reinforcement preparations", providing accurate decision support for operation and maintenance personnel.

[0056] From the data comparison, the present application has significant advantages in evaluation accuracy, response speed and prediction ability. The traditional method usually takes several hours to perform a comprehensive finite element analysis of the tower, while the present application shortens the time to less than 10 minutes through parallel computing; the risk warning changes from "possibility" to specific and time node "predictive" suggestions, with an accuracy of more than 90%. Through the closed-loop feedback mechanism, the accuracy of the model continues to improve during operation, and compared with the initial model, the final comprehensive evaluation error is reduced by about 25%.

[0057] It should be noted that the electrical connection between the above-mentioned units does not necessarily represent the direct connection of the line, and the indirect connection mode can also be applied to the embodiments of the present application as long as the purpose of the present application is achieved. The above-described is only an exemplary embodiment of the present application, and cannot limit the scope of the present application.

[0058] That is, any equivalent changes and modifications made according to the teachings of the present application are still within the scope of the present application. Other embodiments of the present application will be readily apparent to those skilled in the art upon considering the description and practice of the disclosure. The present application is intended to cover any variations, uses or adaptive changes of the present application that follow the general principles of the present application and include common knowledge or conventional technical means in the art not disclosed by the present application.

Claims

1. A computer-aided method and system for assessing the robustness and safety of transmission line towers, characterized in that: Sensor data, environmental data, and historical maintenance data of transmission line towers are acquired and fused to obtain a multi-source dataset; Based on the multi-source dataset, time-series alignment and adaptive filtering of outlier data are performed to generate preprocessed data. Based on the preprocessed data, the mechanical parameters of the finite element analysis are adjusted to generate a dynamically parameterized finite element structure. Simulation calculations are performed based on the dynamically parameterized finite element structure to generate physical index data including stress distribution and deformation trend. The preprocessed data is matched with a preset historical fault database and preset risk rules to generate a preliminary risk index. Based on the environmental data in the preprocessed data, an environmental severity index is determined. The environmental severity index, the physical index data, and the preliminary risk index are combined and analyzed to generate a comprehensive status index. Based on the comprehensive state index, hierarchical feature extraction is performed to separate short-term and long-term features; The short-term and long-term features are fused and adaptively evaluated to output risk warning information.

2. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 1, characterized in that, The generated preprocessed data includes: The sensor data, environmental data, and historical maintenance data in the multi-source dataset are aligned in terms of time and space to generate spatiotemporally aligned data. The environmental context is extracted from the spatiotemporal aligned data, and the filtering threshold is dynamically adjusted based on the environmental context to generate a dynamic filtering threshold. The spatiotemporal aligned data is filtered using the dynamic filtering threshold to remove noise and generate preprocessed data.

3. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 1, characterized in that, The generation of the dynamically parameterized finite element structure includes: The current physical state characterization of the tower structure is comprehensively identified and established from the preprocessed data; Based on the physical state characterization, the material mechanical parameters in the finite element analysis are adaptively calibrated to obtain the calibrated material parameters, which include elastic modulus, Poisson's ratio, yield strength, fatigue damage accumulation factor, and damping characteristics. The current actual environmental load information is extracted and quantified in real time from the preprocessed data. Based on the actual environmental load information, the external load application method, loading point and load size are dynamically determined to obtain the determined external load. By combining the physical state characterization with the actual environmental load information, the boundary conditions and constraints of the finite element analysis are adjusted in a coordinated manner to obtain the adjusted boundary conditions; By combining the calibrated material parameters, the determined external loads, and the adjusted boundary conditions, a dynamically parameterized finite element structure is generated.

4. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 3, characterized in that, The generation of physical index data containing stress distribution and deformation trends includes: The dynamically parameterized finite element structure is divided into multiple computational units according to its physical structure; The computing units are allocated to multi-core processors or cloud computing nodes using a preset task scheduling algorithm to generate the allocated computing tasks. The assigned computational tasks are executed and the computational results are aggregated to generate physical index data.

5. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 1, characterized in that, The generation of the preliminary risk index includes: Extract the current state data from the preprocessed data; The current status data is matched with the historical fault database and the risk rules to generate multiple risk factors; The multiple risk factors are weighted and summed to generate a preliminary risk index.

6. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 3, characterized in that, The generated comprehensive status index includes: Based on the environmental data, feature extraction is performed to obtain wind speed, ice thickness, temperature and humidity, and environmental severity index is calculated. Based on the dynamically parameterized finite element structure, the weighting coefficients of the physical index data are dynamically determined through parameter perturbation analysis. The component-level dynamic safety margin is calculated based on the calibrated material parameters and the physical index data, and the preliminary risk index is normalized and adjusted in combination with the dynamic safety margin to obtain a standardized risk score. The physical index data is weighted using the weighting coefficients and then integrated with the standardized risk score and the environmental severity index to generate a comprehensive status index.

7. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 6, characterized in that, The method further includes: The preliminary risk index is compared with the simulated risk metric value generated based on the physical index data. When the difference comparison result exceeds the preset conflict threshold, a conflict signal is generated. In response to the conflict signal, the parameters of the dynamically parameterized finite element structure are adjusted according to the preset diagnostic logic library to generate optimized finite element parameters; The optimized finite element parameters are used to perform a new simulation calculation to generate updated comprehensive state indices.

8. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 1, characterized in that, The hierarchical feature extraction based on the comprehensive state index, separating short-term and long-term features, includes: Based on the comprehensive state index, a sliding window segmentation is performed to extract the fluctuation and trend features within the window as short-term features. Based on the comprehensive state index, a historical span analysis is performed to extract degradation characteristics and evolution patterns across time scales as long-term features.

9. The computer-aided method for assessing the robustness and safety of transmission line towers according to claim 8, characterized in that, The output risk warning information includes: The short-term features are fused with the long-term features to form a multi-scale state description; By combining the multi-scale state description with the environmental severity index, state evolution is extrapolated to generate risk levels and early warning suggestions for future periods, serving as risk warning information.

10. A computer-aided robustness and safety assessment system for transmission line towers, applied to the computer-aided robustness and safety assessment method for transmission line towers as described in any one of claims 1-9, characterized in that, The system includes: The multi-source data acquisition module is used to acquire sensor data, environmental data, and historical maintenance data of transmission line towers and perform fusion processing to obtain a multi-source dataset. The temporal alignment preprocessing module is used to perform temporal alignment and adaptive filtering of outlier data based on the multi-source dataset to generate preprocessed data. The dynamic finite element modeling module is used to adjust the mechanical parameters of the finite element analysis based on the preprocessed data to generate a dynamic parameterized finite element structure. The simulation calculation and analysis module is used to perform simulation calculations based on the dynamically parameterized finite element structure and generate physical index data including stress distribution and deformation trend. The risk matching and assessment module is used to match the preprocessed data with a preset historical fault database and preset risk rules to generate a preliminary risk index. The comprehensive status calculation module is used to determine an environmental severity index based on the environmental data in the preprocessed data, and to analyze the environmental severity index, the physical index, and the preliminary risk index to generate a comprehensive status index. The hierarchical feature extraction module is used to extract hierarchical features based on the comprehensive state index, separating short-term features and long-term features. The multi-scale risk assessment module is used to perform multi-scale feature fusion and adaptive assessment analysis on the short-term and long-term features, and output risk warning information.

Images