Fatigue life prediction method of pre-stranded type damper for ultra-high voltage ground wire

Abstract

The application discloses a pre-stranded type anti-vibration hammer fatigue life prediction method for extra-high voltage ground wires, and particularly relates to the technical field of power transmission line anti-vibration. Multi-source vibration and environmental data are collected to construct a vibration characteristic time sequence; based on multi-frequency vibration response, amplitude, frequency and strain energy are extracted to establish a multi-frequency collaborative vibration model; structure and material parameters of the anti-vibration hammer are obtained to establish a finite element model and couple with the vibration model to simulate local stress response; a Rainflow counting method and a Miner damage criterion are used to construct a fatigue damage factor distribution matrix; based on damage evolution trend, the residual fatigue life of the anti-vibration hammer is predicted, and the replacement opportunity is evaluated; the method can realize accurate fatigue life prediction and optimal replacement strategy recommendation of the pre-stranded type anti-vibration hammer under complex working conditions, and has high precision, high adaptability and engineering practicability.

Description

Technical Field

[0001] This invention relates to the field of vibration damping technology for power transmission lines, specifically to a method for predicting the fatigue life of pre-stretched vibration dampers used in ultra-high voltage grounding lines. Background Technology

[0002] In ultra-high voltage (UHV) transmission lines, vibration dampers are often installed on conductors or ground wires to suppress fatigue damage caused by wind-induced vibration. Pre-stranded vibration dampers are widely used in UHV ground wires due to their ease of installation, strong grip, and good vibration damping effect. However, because of the special location of the ground wire, it is constantly exposed to complex wind-induced vibration environments, making it prone to multi-frequency vibrations such as micro-vibrations, sub-resonance, and secondary relaxation. This leads to stress concentration at the contact point between the pre-stranded vibration damper and the ground wire, resulting in fatigue cracks and structural degradation.

[0003] Currently, the service performance of pre-stirred vibration dampers largely relies on statistical analysis of field operating years or destructive testing methods. There is a lack of fatigue life prediction models that comprehensively consider factors such as ground wire vibration characteristics, meteorological factors, structural parameters, and material fatigue performance. Traditional methods struggle to accurately predict the fatigue life of pre-stirred vibration dampers under different operating conditions, limiting the optimization of their operation and maintenance strategies. Summary of the Invention

[0004] The purpose of this invention is to provide a method for predicting the fatigue life of a pre-stretched vibration damper for ultra-high voltage grounding wires, in order to address the shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for predicting the fatigue life of a pre-stretched vibration damper for ultra-high voltage grounding wires, comprising:

[0006] Collect multi-source vibration data and corresponding environmental parameter data of the target ground wire during its operating cycle, and construct a vibration characteristic time series.

[0007] Based on the vibration characteristic time series, the amplitude, frequency and strain energy indices of multiple frequency bands are extracted to establish a multi-frequency coordinated vibration model;

[0008] Obtain the structural parameters and material fatigue parameters of the pre-stretched vibration damper, and establish a mechanical finite element model;

[0009] The multi-frequency coordinated vibration model is coupled with the mechanical finite element model to simulate the local stress field distribution of the vibration damper under vibration input, and the fatigue damage factor distribution matrix is ​​obtained.

[0010] Based on the fatigue damage factor distribution matrix and fatigue cumulative damage theory, the fatigue life of the pre-wound vibration damper within the target cycle is calculated.

[0011] Output fatigue life and its distribution law, and evaluate the timing of vibration damper replacement based on damage evolution trend.

[0012] Preferably, the construction of the vibration feature time series includes:

[0013] Vibration response data of the target ground wire under multiple typical meteorological conditions were collected in segments, including sunny weather, gusts, rain, and ice cover.

[0014] Each segment of collected data is preprocessed to extract a multidimensional vibration feature vector, including amplitude, frequency, acceleration, and strain energy.

[0015] The multidimensional feature vectors are mapped to a standardized vibration feature time series to preserve vibration mode evolution information.

[0016] Preferably, the establishment of the multi-frequency coordinated vibration model includes:

[0017] Based on the vibration characteristic time series, the main vibration frequencies of different frequency bands are identified and divided into low-frequency band, mid-frequency band and high-frequency band;

[0018] For each frequency band, a multi-order autoregressive model is used to fit its frequency response function, extract the frequency domain amplitude-frequency characteristics and energy distribution characteristics, and form independent sub-vibration modes;

[0019] Based on the principle of energy superposition and phase coupling conditions, the independent sub-vibration modes are integrated into a unified multi-frequency coordinated vibration model to reflect the vibration coupling effect of the ground wire under different excitation frequencies.

[0020] Preferably, the establishment of the mechanical finite element model includes:

[0021] Based on the structural design drawings of the pre-stranded vibration damper and ground wire, a three-dimensional geometric solid model including the clamping section, conductor section and pre-stranded wire section is constructed, and the contact interface is explicitly modeled.

[0022] Material property parameters, including the elastic modulus, Poisson's ratio, yield strength and fatigue limit of the ground wire and vibration damper components, are applied to the three-dimensional geometric solid model, and friction-contact coupling boundary conditions between the pre-twisted wire and the conductor are introduced.

[0023] An adaptive mesh refinement strategy was adopted to locally refine the contact area and stress concentration area of ​​the pre-twisted wire, establish a finite element analysis model, and set boundary load conditions that match the vibration input.

[0024] Preferably, the coupling of the multi-frequency coordinated vibration model with the mechanical finite element model includes:

[0025] The equivalent excitation parameters of the multi-frequency coordinated vibration model in each frequency band are extracted, including frequency, amplitude and energy per unit time, and a time series input load spectrum is constructed.

[0026] The time-series load spectrum is applied to fixed nodes or regions of the finite element model through a boundary mapping function, which is constructed based on frequency band location, load direction, and node weight coefficients.

[0027] Multi-frequency coupled dynamic response analysis was performed in the finite element platform to obtain the periodic stress response matrix of the key structural region of the vibration damper.

[0028] Preferably, the local stress field distribution of the vibration damper under simulated vibration input, to obtain the fatigue damage factor distribution matrix, includes:

[0029] Based on the finite element results after coupling analysis, the nodal stress response sequence of the pre-stretched vibration damper in multiple vibration cycles is extracted, and a local stress field time distribution map is constructed.

[0030] The nodal stress response sequence was converted into an equivalent cyclic load history, and the stress amplitude and load cycle number were extracted using the Rainflow counting method.

[0031] By combining the material SN curve parameters, the fatigue damage factor of each key node is calculated based on the Miner linear damage accumulation criterion, forming a spatially distributed fatigue damage factor matrix.

[0032] Preferably, the fatigue life of the pre-wound vibration damper within the target cycle is calculated based on the fatigue damage factor distribution matrix and the fatigue cumulative damage theory, including:

[0033] The single-cycle damage value of key nodes or regions is extracted based on the fatigue damage factor distribution matrix, and the equivalent damage rate of the structure per unit time is calculated by combining the ground wire running vibration frequency.

[0034] Based on the importance and load sensitivity of the structural functional area where the node is located, a comprehensive damage growth curve is constructed.

[0035] Combining the target design life cycle and vibration condition statistics, a nonlinear life extrapolation function is used to predict the remaining fatigue life of the overall pre-wound vibration damper, and a life distribution map is output.

[0036] Preferably, the method of assessing the timing of vibration damper replacement based on damage evolution trends includes:

[0037] Based on fatigue life prediction results and historical damage data, a multi-stage evolution curve model containing time-damage relationship is constructed, and the local slope change rate is extracted as the damage growth rate factor.

[0038] The system sets up multi-level replacement criteria, including absolute lifespan threshold, local accelerated damage factor threshold, and regional high-risk distribution judgment conditions, forming a combined replacement early warning mechanism.

[0039] By combining the operation and maintenance cycle with the replacement cost function, an optimization decision model is established, which outputs the optimal replacement window for the vibration damper and a priority ranking table.

[0040] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0041] 1. This invention integrates a multi-frequency coordinated vibration model with a mechanical finite element model to systematically obtain the periodic stress response of a structure under actual service conditions. Combined with a fatigue damage factor distribution matrix, it achieves a spatialized expression of the degree of local fatigue damage. Compared with traditional prediction methods based on static stress or empirical formulas, this invention possesses higher modeling accuracy and environmental adaptability, effectively capturing the multi-frequency coupled response behavior of ground wires under complex wind-induced vibrations, thus improving the scientific rigor and accuracy of fatigue life assessment.

[0042] 2. This invention introduces damage evolution trend analysis and multi-level replacement early warning criteria to construct an optimal replacement timing assessment mechanism based on life prediction results. This supports hierarchical management of structural functions and priority ranking of vibration damper replacement, significantly improving the initiative and economy of operation and maintenance planning. While ensuring safety redundancy, this method also considers actual operation and maintenance constraints and failure risk control, making it suitable for engineering scenarios with extremely high structural reliability requirements, such as ultra-high voltage transmission lines. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0044] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] For examples, please refer to Figure 1 As shown in this embodiment, the fatigue life prediction method for pre-stretched vibration dampers used in ultra-high voltage grounding wires includes:

[0047] Collect multi-source vibration data and corresponding environmental parameter data of the target ground wire during its operating cycle, and construct a vibration characteristic time series.

[0048] Based on the vibration characteristic time series, the amplitude, frequency and strain energy indices of multiple frequency bands are extracted to establish a multi-frequency coordinated vibration model;

[0049] Obtain the structural parameters and material fatigue parameters of the pre-stretched vibration damper, and establish a mechanical finite element model;

[0050] The multi-frequency coordinated vibration model is coupled with the mechanical finite element model to simulate the local stress field distribution of the vibration damper under vibration input, and the fatigue damage factor distribution matrix is ​​obtained.

[0051] Based on the fatigue damage factor distribution matrix and fatigue cumulative damage theory, the fatigue life of the pre-wound vibration damper within the target cycle is calculated.

[0052] Output fatigue life and its distribution law, and evaluate the timing of vibration damper replacement based on damage evolution trend.

[0053] In this invention, to achieve accurate prediction of the fatigue life of the pre-stirred vibration damper, it is first necessary to construct a vibration characteristic time series that can truly reflect the multi-source excitation characteristics and response features during the service process of the ground wire. The construction steps of the vibration characteristic time series are not only the basis for subsequent mechanical modeling and damage analysis, but also directly determine the accuracy and adaptability of the life prediction model.

[0054] First, vibration response data of the target ground wire under multiple typical meteorological conditions are collected in segments. Since the stress state and vibration behavior of the ground wire vary significantly under different meteorological conditions, this invention selects representative and widely covered typical meteorological conditions, specifically including four conditions: sunny, gusty, rainy, and icing. In practical applications, historical operating data and meteorological data of the target transmission line can be classified and filtered to establish a corresponding meteorological-operating condition labeling system. Furthermore, the continuous vibration data obtained from sensor monitoring is divided by operating condition and sliced ​​into time periods, thereby forming a structured multi-condition vibration dataset.

[0055] Secondly, the collected ground wire vibration response data for each segment is preprocessed, and key multidimensional vibration feature vectors are extracted. To reduce noise interference in the raw data and highlight key dynamic features, this invention preferably employs methods such as filtering and denoising, normalization, and sliding window segmentation during data preprocessing. Subsequently, feature extraction algorithms are used to calculate multiple physically meaningful vibration feature indices for each data segment. These indices include, but are not limited to, the following four: vibration amplitude, frequency, acceleration, and strain energy. Among them, vibration amplitude is used to characterize the maximum displacement response of the ground wire, frequency reflects the dominant mode of vibration, acceleration is an important indicator of vibration intensity, and strain energy comprehensively reflects the external energy absorbed by the structure and is closely related to fatigue damage. The above multidimensional vibration features together constitute a feature vector to describe the comprehensive vibration state of the ground wire in each time period.

[0056] Furthermore, the aforementioned multidimensional vibration feature vectors are mapped into a standardized vibration feature time series to preserve the evolution trend and variation law of the vibration mode throughout the entire operating cycle. In this invention, feature normalization and time alignment are used to uniformly standardize feature data from different time periods and dimensions, making them scalable for analysis. Subsequently, combined with timestamp information, all feature vectors are reconstructed according to the sampling order to form a vibration feature matrix arranged in ascending order of time. Each row of this matrix represents the vibration state at a certain moment, and each column corresponds to a certain type of feature value. Further, algorithms such as interpolation completion and window sliding can be used to ensure the continuity and smoothness of the time series, thereby effectively describing the internal response variation trend of the ground wire under different meteorological conditions caused by external excitation. This vibration feature time series provides a data foundation for subsequent mechanical simulation analysis, fatigue damage factor calculation, and life prediction.

[0057] In order to accurately characterize the multi-frequency response behavior of the ground wire under wind-induced excitation of different frequencies during actual operation, and to further provide input basis for the fatigue life prediction of the pre-stretched vibration damper, this invention proposes a method for constructing a multi-frequency collaborative vibration model that integrates the response characteristics of multiple main vibration frequency bands, based on the completed vibration characteristic time series.

[0058] First, based on the constructed vibration characteristic time series, spectral analysis is used to perform Fourier transform or wavelet transform processing on the ground wire vibration data at different time periods to identify the dominant energy frequency. Specifically, power spectral density analysis can be performed on the acceleration or displacement signal within each time window, and the range of the dominant frequency can be determined by finding the frequency point corresponding to the energy peak.

[0059] To facilitate subsequent modeling and feature summarization, this invention divides the dominant frequency into three intervals according to its frequency range:

[0060] Low frequency band: frequencies less than 10 Hz, mainly corresponding to low-amplitude, large-period vibration behavior under light wind conditions;

[0061] Mid-frequency band: The frequency is between 10 Hz and 30 Hz, and it is generally in a typical sub-resonant state;

[0062] High frequency band: frequencies above 30 Hz, commonly seen in high-frequency local responses during gust disturbances or ice shedding processes.

[0063] The above frequency band allocation is based on the statistical results of a large amount of wind-induced vibration monitoring data and has good engineering adaptability and generalization ability.

[0064] After dividing the frequency bands, a multi-order autoregressive model (AR model) was used to model the vibration data in each band in the frequency domain, extracting its frequency response characteristics and energy distribution patterns. This model can effectively fit the dynamic behavior of the ground wire at different frequencies and capture its evolution trend over time.

[0065] The specific modeling steps are as follows:

[0066] The vibration time series within each frequency band is stabilized to eliminate trend terms and non-stationary disturbances;

[0067] Select an appropriate model order based on information criteria (such as the Akaike Information Criterion or the Bayesian Criterion);

[0068] Fit the coefficients of the autoregressive model to construct the frequency response function for this frequency band;

[0069] The amplitude-frequency characteristic curve and energy distribution per unit time of this frequency band are further calculated from the frequency response function.

[0070] Taking acceleration response as an example, if the autoregressive model output in a certain frequency band is in the form of a linear difference equation, its frequency response can be obtained through system function transformation. Then, the average value per unit energy in that frequency band is calculated, and the amplitude and energy are mapped to the response characteristics of that frequency band to form an independent sub-vibration mode.

[0071] Each sub-vibration mode is described by the main response frequency of that frequency band, the corresponding amplitude-frequency response characteristics, the unit energy index, and the model parameters, forming a basic unit that can be used for structural coupling.

[0072] To more realistically reflect the collaborative response behavior of the actual ground wire under multi-frequency excitation, this invention further proposes to couple and integrate the above-mentioned multiple sub-vibration modes to construct a unified multi-frequency collaborative vibration model.

[0073] The following two principles are adopted in the integration process:

[0074] Energy superposition principle: The energy per unit time of each frequency band mode is normalized and weighted according to its proportion to form the overall energy input sequence. The energy weights can be set based on the normalized values ​​of the peak power spectral density of each frequency band;

[0075] Phase coupling condition: For submodes with neighboring frequency distributions, their phase relationships are analyzed, and the phase superposition method is adjusted through a coupling matrix to avoid frequency interference or spurious enhancement. This matrix is ​​constructed from the frequency spacing and phase delay function, controlling the response synchronization between modes.

[0076] Finally, the sub-modes are spliced ​​and superimposed along the time axis to form a multi-dimensional state sequence representing the actual response process, which is the multi-frequency coordinated vibration model. This model not only preserves the independent response characteristics of each frequency band, but also reflects the energy interaction and structural coupling effects that may occur between frequency bands during actual service.

[0077] The model can be visualized as a three-dimensional vibration response spectrum, where the horizontal axis represents time, the vertical axis represents frequency, and the amplitude or energy is the third dimension, which facilitates analysis and comparison.

[0078] To achieve accurate analysis of stress distribution and fatigue damage behavior of pre-stressed vibration dampers under service conditions, this invention proposes a complete mechanical finite element modeling method based on obtaining the structural parameters and material fatigue parameters of the vibration damper.

[0079] First, a three-dimensional geometric solid model is constructed, and the contact interface is explicitly modeled. This step begins by constructing a three-dimensional geometric solid model based on the structural design drawings of the pre-stretched vibration damper and its grounding wire. The pre-stretched vibration damper typically consists of a clamping section, a pre-stretched wire section, and a connecting conductor, with each component exhibiting a complex spiral structure and relative nesting relationship in space.

[0080] During the modeling process, to accurately reproduce the actual structure of the vibration damper, CAD or CAE modeling software (such as SolidWorks, Creo, or ANSYS DesignModeler) should be used to construct a complete three-dimensional structural model based on the original two-dimensional engineering drawings. In particular:

[0081] The clamping section should include an arc-shaped cavity and a bolt preload system for actually clamping the ground wire;

[0082] The conductor segment should take into account the diameter, nominal tensile strength, and outer coating of the ground wire;

[0083] The pre-twisted section should construct a spiral path entity to accurately reproduce the pitch, wrap angle, number of strands, and winding method of each pre-twisted wire;

[0084] The contact interface needs to be modeled explicitly, and the contact area between the outer surface of the pre-twisted wire and the surface of the conductor should be defined separately to ensure that the contact stress and slip behavior can be accurately calculated in the future.

[0085] Through the above modeling strategies, a complete high-fidelity three-dimensional geometric model is formed, which has the ability to realistically reflect the details of the stressed parts.

[0086] After completing the geometric modeling, actual material property parameters need to be applied to each part of the model. Material properties can be obtained based on the manufacturer's technical specifications or through experimental testing. Key parameters include:

[0087] Elastic modulus (used to describe the stress-strain relationship of a material in the elastic stage).

[0088] Poisson's ratio (reflects the ratio of transverse strain to longitudinal strain in a material).

[0089] Yield strength (the boundary used to distinguish between elastic and plastic deformation);

[0090] Fatigue limit (describes the load-bearing capacity of a material under cyclic loading).

[0091] For example, a typical steel vibration damper has an elastic modulus of about 2.0 × 10¹¹ Pa, a Poisson's ratio of 0.3, a yield strength of over 400 MPa, and a fatigue limit that is generally in the range of 200 to 300 MPa depending on the working conditions.

[0092] More importantly, this invention performs friction-contact coupling modeling of the contact interface between the pre-twisted wire and the conductor. Using a "surface-to-surface" contact definition method, the following parameters are set in finite element platforms such as ANSYS and ABAQUS:

[0093] The coefficient of friction is set to 0.3~0.5, and the specific value is adjusted according to the coating type and the clamping force.

[0094] Normal contact is achieved by applying constraints using the penalty function method to simulate the pre-twisted wire clamping force;

[0095] The contact type is set to non-linear hard contact, which supports relative slippage between nodes but does not allow penetration.

[0096] This coupling mechanism can effectively reflect the micro-slippage, nonlinear load transfer and stress concentration phenomena that occur in the pre-twisted wire during the periodic vibration of the ground wire, and is the core prerequisite for realizing high-precision fatigue analysis.

[0097] To improve the spatial resolution of stress analysis and reduce computational resource consumption, this invention adopts a local mesh adaptive refinement strategy when establishing the finite element analysis model.

[0098] The specific steps are as follows:

[0099] The mesh is refined in the pre-twisted wire contact area, the inner wall of the clamping section, and the structural connection transition area, with the unit size controlled between 0.1 and 0.5 mm;

[0100] The remaining areas use larger dimensions (e.g., 2 mm or more) to reduce the overall model size;

[0101] Tetrahedral structural elements (such as C3D10 elements) are preferred for mesh type to accommodate complex geometries;

[0102] Apply node compatibility constraints between contact surfaces to avoid local stiffness imbalance.

[0103] Furthermore, to simulate the stress on the ground wire under actual vibration excitation, boundary load conditions matching those of the multi-frequency vibration model need to be applied to the model. These mainly include:

[0104] Apply fixed constraints or simply supported conditions to both ends of the ground wire to simulate its fixed state at the tower head;

[0105] A periodic transverse load is applied to the vibration loading zone, and the frequency and amplitude of the load are determined based on the output of the aforementioned multi-frequency cooperative model.

[0106] Preload simulation simulates the bolt clamping state by applying axial constraints or equivalent preload loads to the clamping parts.

[0107] These load boundary settings ensure consistency between external inputs and actual operating conditions during the simulation process, enabling accurate prediction of the dynamic response within the structure.

[0108] To accurately predict the structural fatigue response and damage distribution of pre-stirred vibration dampers under actual vibration conditions, this invention proposes a highly coupled data-driven modeling method based on the construction of a multi-frequency collaborative vibration model and finite element modeling, thereby enabling the exchange of mechanical information and dynamic response mapping between the two models.

[0109] To ensure that the external excitation characteristics obtained from the vibration model can be used as effective boundary conditions input into the finite element simulation platform, three key steps need to be completed:

[0110] First, equivalent excitation characteristics of ground wire vibration in different frequency bands are extracted from the multi-frequency coordinated vibration model, including dominant frequency, vibration amplitude, and vibration energy per unit time. These parameters are derived from the frequency domain analysis output of the vibration characteristic time series and are usually represented as a combination of multiple frequency components, each of which has a specific amplitude and energy proportion.

[0111] To achieve dynamic load input, the frequency domain excitation parameters mentioned above need to be converted into a time domain load spectrum. Let the dominant frequency of the i-th frequency band be fi, the amplitude be Ai, and the energy per unit time be Ei. Then its equivalent excitation function can be expressed as: Excitation function: Where φi is the initial phase of the frequency band. The total excitation function for multiple frequency bands can be constructed by energy-weighted superposition: Total excitation function: ,in , representing the energy normalization weight. This function generates a set of equally spaced sampling points in the discrete-time dimension, which constitute the time-series input load spectrum and serve as the input driver for the coupled solution.

[0112] Applying the constructed time-series load spectrum to the finite element model requires addressing three key issues: load location, direction, and application method. This invention proposes a boundary mapping function based on node distribution to achieve accurate projection of frequency band excitation onto structural node loads.

[0113] Let the load application region be Ω, and the node set be... Where k is the total number of nodes, the load input of each node is described by the following function: Where: αj is the node position weighting coefficient, generated based on the distance function from the node to the load center; Dj is the direction vector, defining the load application direction (e.g., perpendicular to the ground wire); F(t) is the total excitation function. This function ensures that the load is reasonably distributed on the structure, conforming to the characteristics of wind-induced or linear vibration experienced by the actual vibration damper.

[0114] After inputting the aforementioned boundary excitations into the finite element model, a transient dynamic analysis is performed in a structural analysis platform (such as ANSYS Workbench, ABAQUS, or MSC.Nastran). The analysis process is driven by vibration excitation, and the system automatically calculates the stress-strain state of the structure at each time step, ultimately outputting the periodic stress response matrix of the key structural regions.

[0115] The matrix takes the following form:

[0116] S(i, t) represents the principal stress or equivalent stress value of the i-th structural node at time t.

[0117] This stress matrix provides an accurate basis for cyclic stress inputs for subsequent fatigue analysis.

[0118] After obtaining the periodic stress response results, they need to be further converted into the load spectrum required for fatigue analysis, and then life damage calculations are performed in conjunction with material fatigue performance parameters. This involves the following three steps:

[0119] By analyzing the time-series stress response data of key nodes, a spatial-temporal distribution map is constructed. For each key region node, the following structure is formed:

[0120] σj(t): The stress time series at node j;

[0121] By combining finite element geometric information, "local stress field animation" or "nodal stress cloud map sequence" can be generated, reflecting the dynamic stress change trend and concentration area caused by local vibration.

[0122] The stress time series σj(t) of each node is input into the Rainflow Counting Algorithm to extract the amplitude and number of various cyclic stresses. This algorithm is a standard method for fatigue load analysis, capable of decomposing irregular stress waveforms into a series of regular closed loops.

[0123] Each node will output a set of cyclic load records: Where σa represents the stress amplitude, N is the corresponding cycle number, and the load spectrum of the node is composed of σa.

[0124] Obtain the SN curve of the vibration damper material, i.e., the logarithmic relationship between stress amplitude and cycle life: Where C and m are material constants, and Nf is the tolerable number of cycles at stress amplitude σa. The total fatigue damage factor Dj of the node is calculated using the Miner linear damage accumulation criterion. Traverse all critical structural nodes to form a spatially distributed fatigue damage matrix: This matrix can be visualized as a three-dimensional damage heatmap to identify high-risk damage concentration areas and guide structural design optimization or maintenance strategies.

[0125] After obtaining the fatigue damage factor distribution matrix of the pre-wound vibration damper, in order to achieve a quantitative assessment of the structural life, this invention further proposes a life prediction method that combines fatigue cumulative damage theory, key area identification and life nonlinear extrapolation. This method can accurately calculate the remaining fatigue life of the vibration damper within the target operating cycle and has the ability to visualize the life.

[0126] First, based on the fatigue damage factor distribution matrix constructed in the previous steps, nodes or regions with significant damage values ​​are selected as the main control points for fatigue life analysis. The stress cyclic load borne by each node j in a complete vibration loading cycle has been extracted using the Rainflow counting method and matched with the material's SN curve to obtain its single-cycle fatigue damage value, denoted as dj.

[0127] Considering that the structure operates under continuous vibration, damage accumulates linearly or nonlinearly over time. Therefore, given the vibration frequency fv (unit: Hertz, i.e., period / second), the equivalent damage rate rj of the structure within a unit time t=1 second can be expressed as: This value represents the rate at which the damage increases at this node per second. By summarizing the damage rate data of key nodes j = 1, 2, ..., n, a preliminary multi-point damage input for the entire structure is constructed.

[0128] To improve the representativeness and sensitivity of the prediction model, this step also introduces a "key node screening mechanism". For example, a threshold dthresh is set, and only nodes with dj≥dthresh are modeled for subsequent lifetimes to reduce computational redundancy.

[0129] Pre-stretched vibration dampers typically consist of multiple functional areas, including a clamping area, a pre-stretched wire wrapping area, and a conductor contact area. Each area exhibits significantly different load response characteristics and structural importance during service. Therefore, to achieve more accurate life assessment, this invention innovatively introduces a structural functional area weighting mechanism to construct a comprehensive damage growth curve.

[0130] The specific method is as follows:

[0131] The structure is divided into m regions according to function, denoted as m. ;

[0132] For each functional region Ωi, a weighting coefficient βi is assigned. This coefficient can be determined by weighting based on the following factors:

[0133] Load concentration (e.g., localized stress peaks);

[0134] Fatigue sensitivity (the influence of material and structural morphology on fatigue response);

[0135] Importance of the project (e.g., the severity of the consequences of failure);

[0136] The damage rate rΩi for each region is calculated as follows: Where ni is the number of critical nodes contained in region Ωi;

[0137] The final comprehensive damage growth curve R(t) is in the form of a weighted superposition: This curve represents the weighted cumulative damage growth of the entire structure per unit time under actual vibration loading, and has structural characteristic sensitivity and engineering adaptability.

[0138] Considering that the vibration input experienced by vibration dampers under actual working conditions is not completely constant over different time periods, but fluctuates with weather changes, service life, and operating mode, using a linear cumulative criterion for life extrapolation may underestimate or overestimate the remaining life. Therefore, this invention introduces a nonlinear life extrapolation function to achieve a life prediction that is closer to engineering reality, while taking into account statistical vibration intensity changes.

[0139] This model is based on the comprehensive damage growth curve R(t), combined with the actual monitoring and statistical analysis of the operating condition fluctuation factor γ(t) (reflecting the relative load intensity change per unit time), and defines the instantaneous damage rate function r(t): γ(t) can be determined based on historical wind speed data, typical operating condition distribution, etc. For example: during normal operation: γ(t) ≈ 1.0; during storm or icing periods: γ(t) > 1.0; during periods of low vibration: γ(t) < 1.0;

[0140] Based on this, the instantaneous damage rate is integrated over time to obtain the total cumulative damage Dtotal(T): According to Miner's linear damage criterion, the structure reaches its fatigue limit when Dtotal(T) reaches 1. Therefore, solving the integral equation inversely yields the remaining fatigue life (Tremaining): To facilitate operational and maintenance decisions, this invention further visualizes the Tremaining of each key area or node as a spatial heat map, forming a fatigue life distribution map. This map reflects the expected remaining life of the structure in different areas, supporting the formulation of strategies for local replacement, reinforcement, or periodic inspection.

[0141] After completing the modeling and prediction of the fatigue life of the pre-wound vibration damper, this invention further proposes a replacement timing assessment method based on damage evolution trend. By outputting the structural fatigue life distribution law, constructing a multi-dimensional damage trend model, setting multi-level early warning criteria, and optimizing the replacement timing, predictive maintenance and hierarchical management of the vibration damper throughout its entire life cycle can be achieved.

[0142] In the aforementioned fatigue life prediction module, the remaining fatigue life value Tremain(j) of all key nodes in the finite element model of the pre-wound vibration damper has been calculated, forming a fatigue life distribution matrix: T = { Tremain(1), Tremain(2), ..., Tremain(n)}; this life distribution matrix represents the expected remaining life of each part of the structure under the current load and operating conditions, exhibiting significant spatial differences and functional area concentration. To improve its engineering interpretability, this invention performs the following visualization processing and pattern extraction on the life distribution:

[0143] Lifetime distribution heatmap output: Based on the three-dimensional geometric model of the structure, the Tremain of each node is mapped in the form of color gradation to construct a fatigue life heatmap, which intuitively presents high-risk (short life) areas.

[0144] Lifetime statistical histogram: Divide the node lifetime into intervals (e.g., with 50-hour intervals), draw a lifetime distribution frequency map, and determine the lifetime concentration intervals and distribution skewness.

[0145] Regional lifetime mean and standard deviation analysis: Taking structural functional regions as units, the lifetime mean and dispersion of each region are statistically analyzed to identify key structural segments with large performance fluctuations;

[0146] High-risk node screening table: Set a lifespan threshold Tthresh, extract information such as node number, location, and stress level for nodes with lifespans below this value, and generate a list of high-risk locations.

[0147] This output not only provides basic data for assessing replacement timing, but also supports operations and maintenance personnel in quickly locating risk areas.

[0148] To avoid sudden structural failures and extend the safe service life of equipment, this invention proposes a replacement timing assessment strategy based on damage evolution trends, in addition to fatigue life output.

[0149] First, the cumulative fatigue damage Dj(t) at key nodes is recorded or predicted over multiple operating cycles, forming a multi-stage evolution curve that changes over time. To characterize the damage evolution process, a piecewise linear fitting and local curvature recognition algorithm is used to obtain the growth rate rj(t) for each stage: Where Δt is the width of adjacent time intervals; simultaneously, a local rate of change δj(t) is introduced to represent the trend of damage growth: This factor is used to determine whether a node has entered the fatigue acceleration stage. If δj(t) > δthresh, it is considered to be in an accelerated growth phase, triggering a replacement warning.

[0150] This invention innovatively employs a multi-dimensional criterion combination strategy, comprehensively considering absolute lifetime value, growth trend, and spatial clustering effect, to construct the following three types of replacement triggering conditions:

[0151] If the node lifetime Tremain(j) ≤ Tmin, it enters a level 1 warning state and it is recommended to replace it first.

[0152] If there exists δj(t)≥δthresh and rj(t) continues to rise for more than the preset time period Δtcrit, it is determined to be an accelerated deterioration segment and enters the level 2 warning state;

[0153] If more than θ% of nodes in a certain functional area enter a warning state (Level 1 or Level 2), the entire area is marked as a high-risk area, and regional maintenance or replacement is recommended.

[0154] It should be noted that the threshold θ is set based on existing experience in high-voltage transmission structure health assessment and simulation data analysis. For example, the basic recommended range of θ values ​​is as follows:

[0155] θ∈[10%, 30%], where: θ≈10%: used for high-sensitivity monitoring of key functional areas (such as clamping sections); θ≈20%: suitable for routine risk identification of most structural areas; θ≈30%: suitable for areas with complex structures but high fault tolerance, reducing the risk of false alarms.

[0156] The combination of three criteria constitutes a complete multi-level early warning mechanism, which has the ability to identify three dimensions of time, rate and space, and can be flexibly adjusted according to the structural characteristics of the vibration damper.

[0157] Considering that replacing vibration dampers involves actual operation and maintenance costs such as transportation, operational risks, and power outage coordination, this invention further introduces a replacement timing optimization model after outputting the replacement criteria, in order to achieve a balance between life risk and economy.

[0158] Construct the cost function Ctotal(i), which consists of the following two parts: Ctotal(i) = Cr(i) + Cf(i); where: Cr(i) is the replacement cost, which is related to location, working conditions, batch operation, etc.; Cf(i) is the potential failure cost, which is proportional to the probability of failure risk and the degree of failure impact of the structure.

[0159] Provided that Tremain(i) ≥ Tmargin (safety margin), the node or region with the smallest Ctotal(i) is selected for replacement.

[0160] Using dynamic programming or greedy search algorithms, the following information can be output:

[0161] We recommend changing the time window (e.g., expecting it to be from month X to month Y);

[0162] Priority ranking table for replacing high-risk structural components;

[0163] A list of collaborative task suggestions (based on geographic location and task window aggregation).

[0164] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting the fatigue life of a pre-stranded type damper for extra-high voltage ground wires, characterized by: Comprise: Collecting multi-source vibration data and corresponding environmental parameter data of the target ground wire in a running cycle, and constructing a vibration feature time sequence, including: Segmenting and collecting vibration response data of the target ground wire under multiple typical meteorological conditions, including sunny day, gust, rainfall, and ice cover; Preprocessing each segment of collected data, and extracting multi-dimensional vibration feature vectors including amplitude, frequency, acceleration, and strain energy; Mapping the multi-dimensional vibration feature vectors to a standardized vibration feature time sequence to retain vibration mode evolution information; Based on the vibration feature time sequence, extracting amplitude, frequency, and strain energy indicators of multiple frequency bands, and establishing a multi-frequency collaborative vibration model, including: Identifying the main vibration frequencies of different frequency bands according to the vibration feature time sequence, and dividing them into low, medium, and high frequency bands; For each frequency band, use a multi-order autoregressive model to fit its frequency response function, extract frequency domain amplitude-frequency characteristics and energy distribution characteristics, and form independent sub-vibration modes; Based on the energy superposition principle and phase coupling conditions, fuse each independent sub-vibration mode into a unified multi-frequency collaborative vibration model to reflect the vibration coupling effect of the ground wire under different excitation frequencies; Obtaining the structure parameters and material fatigue parameters of the pre-stranded anti-vibration hammer, and establishing a mechanical finite element model; Coupling the multi-frequency collaborative vibration model with the mechanical finite element model to simulate the local stress field distribution of the anti-vibration hammer under vibration input, and obtaining a fatigue damage factor distribution matrix, including: Extracting equivalent excitation parameters of the multi-frequency collaborative vibration model in each frequency band, including frequency, amplitude, and energy per unit time, and constructing a time sequence input load spectrum; Applying the time sequence input load spectrum to the fixed nodes or regions of the finite element model through a boundary mapping function, which is constructed based on frequency band position, load direction, and node weight coefficient; Performing multi-frequency coupled dynamic response analysis in the finite element platform to obtain a periodic stress response matrix of the key structure region of the anti-vibration hammer, including: Based on the finite element results after coupling analysis, extracting the node stress response sequence of the pre-stranded anti-vibration hammer in multiple vibration cycles, and constructing a local stress field time distribution map; Converting the node stress response sequence to an equivalent cyclic load history, and extracting the stress amplitude and load cycle number using the Rainflow counting method; Combining the material S-N curve parameters, calculating the fatigue damage factor of each key node based on the Miner linear damage accumulation criterion, and forming a spatially distributed fatigue damage factor matrix; Based on the fatigue damage factor distribution matrix and fatigue cumulative damage theory, calculating the fatigue life of the pre-stranded anti-vibration hammer in the target cycle, including: Based on the fatigue damage factor distribution matrix, extracting the single-cycle damage value of the key nodes or regions, and combining the ground wire running vibration frequency to calculate the equivalent damage rate of the structure per unit time; According to the importance and load sensitivity of the structure function area where the node is located, constructing a comprehensive damage growth curve; Combining the target design life cycle and vibration condition statistical data, using a nonlinear life deduction function to predict the remaining fatigue life value of the overall pre-stranded anti-vibration hammer, and outputting a life distribution map; Output fatigue life and its distribution law, and evaluate the replacement timing of the damper according to the damage evolution trend.

2. The pre-stranded type damper fatigue life prediction method for UHV ground wires according to claim 1, characterized in that: The method comprises the following steps: The method comprises the following steps: Based on the structural design drawings of the pre-stranded damper and the ground wire, a three-dimensional geometric entity model including the clamping section, the wire section and the pre-stranded wire section is constructed, and the contact interface is explicitly modeled; Material property parameters are applied to the three-dimensional geometric entity model, including the elastic modulus, Poisson's ratio, yield strength and fatigue limit of the ground wire and the damper components, and the friction-contact coupling boundary conditions between the pre-stranded wire and the wire are introduced; 3. The pre-stranded type damper fatigue life prediction method for UHV ground wires according to claim 1, characterized in that: The local encryption division of the pre-stranded wire contact area and the stress concentration area is carried out by using the adaptive mesh encryption strategy, the finite element analysis model is established, and the boundary load conditions matched with the vibration input are set. The method comprises the following steps: Based on the fatigue life prediction results and the historical damage data, a multi-stage evolution curve model containing time-damage relationship is constructed, and the local slope change rate is extracted as the damage growth rate factor; A multi-level replacement criterion is set, including an absolute life threshold, a local accelerated damage factor threshold and a regional high-risk distribution judgment condition, forming a combined replacement early warning mechanism; An optimization decision model is established by combining the operation and maintenance cycle with the replacement cost function, and the best replacement timing window and priority ranking table of the damper are output.

Images