Iced conductor galloping amplitude prediction method and system, and storage medium

Abstract

The invention discloses an iced conductor galloping amplitude prediction method and system and a storage medium, and belongs to the technical field of power system safety monitoring and prediction.The method comprises the steps that environmental meteorological data of a power transmission conductor are collected in real time and preprocessed; inputting the preprocessed environmental meteorological data into a pre-constructed conductor dynamic response prediction model under the normal working condition to obtain a galloping amplitude prediction value of the transmission conductor under the normal working condition; inputting the preprocessed environmental meteorological data into a pre-constructed icing thickness prediction model to obtain an icing thickness prediction value; determining a lift coefficient and a resistance coefficient based on the preprocessed wind speed data and the icing thickness prediction value; and calculating an equivalent aerodynamic coefficient according to the lift coefficient and the resistance coefficient, converting the galloping amplitude predicted value of the transmission conductor under the normal working condition in combination with a pre-constructed nonlinear conversion formula, obtaining the galloping amplitude predicted value of the transmission conductor under the current icing condition, and transmitting the galloping amplitude predicted value to a monitoring center in real time.

Description

Technical Field

[0001] The present invention relates to power system safety monitoring and prediction technology, and in particular to a method, system and storage medium for predicting the galloping amplitude of ice-covered conductors by using a fully connected neural network combined with real-time wind speed, ambient temperature and humidity, ice thickness and other parameters. Background Art

[0002] Under extreme winter weather conditions such as low temperatures, strong winds, and heavy rainfall, transmission lines are often threatened by icing. Icing not only significantly increases the weight of conductors but also alters the stresses on them, causing them to vibrate and sway abnormally under wind pressure. This can lead to serious accidents such as line breakage and tower collapse, posing a significant risk to the stable operation of the power grid and the security of power supply. Particularly in northern my country and high-latitude regions, where severe weather conditions are frequent, icing is a particularly prominent issue, posing a significant challenge to power system maintenance and dispatch. Given this reality, preventing and promptly addressing the risk of line swaying caused by icing has become a critical issue and a pressing challenge in power grid safety management.

[0003] Prior art document 1 (CN117272753A) discloses a solution in the field of transmission line ice dance warning technology. It uses gray correlation analysis to screen meteorological parameter characteristics, combines a support vector machine model to predict ice thickness, and generates a warning signal based on the ice degree and meteorological data. Its technical approach focuses on the static prediction of ice thickness and the correlation analysis of meteorological parameters. As a result, the warning results are insufficient to quantitatively describe the actual dynamic response of the conductor under ice conditions. In addition, the analysis of the impact of meteorological parameter changes on the dynamic response during the ice growth, maintenance, and melting stages lacks refined processing, limiting the real-time and accuracy of the warning.

[0004] Prior Art Document 2 (CN116822933A) discloses a solution for assessing the risk of transmission line galloping damage. This solution extracts future icing trend curves through cluster analysis of historical icing data, and generates galloping intensity curves based on line structural parameters to assess risk levels. This technical approach relies on statistical analysis of historical data and static parameter mapping, lacking the ability to quantitatively correct real-time dynamic responses. Furthermore, it lacks an effective mechanism for addressing the nonlinear mutations that occur during icing evolution under complex meteorological conditions. Consequently, the accuracy and real-time performance of risk assessment results are insufficient to meet practical requirements. Summary of the Invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a method, system and storage medium for predicting the galloping amplitude of an ice-covered conductor.

[0006] The present invention adopts the following technical solutions.

[0007] A first aspect of the present invention provides a method for predicting the galloping amplitude of an ice-covered conductor, comprising the following steps:

[0008] Real-time collection and pre-processing of environmental meteorological data of transmission lines, including wind speed, temperature, humidity and precipitation;

[0009] The pre-processed environmental meteorological data is input into the pre-built conductor dynamic response prediction model under normal working conditions to obtain the predicted value of the transmission line's galloping amplitude under normal working conditions;

[0010] The pre-processed environmental meteorological data is input into the pre-built ice thickness prediction model to obtain the ice thickness prediction value;

[0011] Based on the pre-processed wind speed data and the predicted ice thickness, the lift coefficient and drag coefficient are determined according to the pre-established mapping relationship between ice thickness and aerodynamic coefficients.

[0012] The equivalent aerodynamic coefficient is calculated based on the lift coefficient and drag coefficient. The predicted value of the transmission line's dancing amplitude under normal operating conditions is converted using a pre-built nonlinear conversion formula. The predicted value of the transmission line's dancing amplitude under current icing conditions is obtained and transmitted to the monitoring center in real time.

[0013] Optionally, the construction of the conductor dynamic response prediction model under normal operating conditions includes:

[0014] Under non-icing conditions, with wind speed, temperature, humidity and inherent parameters of conductors as input features and conductor dancing amplitude as output target, a dynamic response mapping framework of input features and output targets is established based on historical monitoring data. A fully connected neural network is used to construct a nonlinear mapping model, and the network weight parameters are optimized through the mean square error loss function to achieve functional relationship fitting between input features and dancing amplitude.

[0015] Optionally, the construction of the ice thickness prediction model includes:

[0016] Temperature, humidity, precipitation, and liquid water content were selected as input features, and the thickness of ice on the conductor surface was used as the output target. A mapping relationship framework between meteorological parameters and ice thickness was established based on the physical mechanism of icing.

[0017] A deep learning regression model is used to train model parameters through icing test data to achieve nonlinear fitting of the relationship between environmental meteorological conditions and ice thickness.

[0018] Optionally, establishing the mapping relationship between ice thickness and aerodynamic coefficient includes:

[0019] Three-dimensional modeling of the ice-covered conductor was performed using computational fluid dynamics software, and aerodynamic simulations were carried out under different ice thicknesses, ice types, and wind speeds to obtain a data set related to the lift coefficient, drag coefficient, ice thickness, and wind speed. Based on the data set, a six-variable polynomial empirical relationship model of the lift coefficient and drag coefficient was established. Its expression consists of a linear combination of a constant term, a linear / quadratic term of ice thickness, a linear / quadratic term of wind speed, and an ice thickness-wind speed interaction term.

[0020] Optionally, performing aerodynamic simulation using computational fluid dynamics software includes:

[0021] Based on the conductor geometric parameters, different ice thicknesses and different ice types, a three-dimensional geometric model of the conductor is constructed and the model is meshed.

[0022] Set the simulation boundary conditions, including the incoming wind speed, incoming flow direction and turbulence model, and define the conductor surface as a solid wall with no slip;

[0023] Configure solution parameters, including time step and convergence threshold, and complete simulation environment initialization;

[0024] In the simulation environment, the Navier-Stokes equations are solved by the fluid module to obtain instantaneous aerodynamic data on the surface of the ice-covered conductor;

[0025] The instantaneous aerodynamic data output by the fluid module is transferred to the structure module, and the lift coefficient and drag coefficient are calculated based on the material properties of the wire.

[0026] Optionally, the performing of aerodynamic simulation using computational fluid dynamics software further includes:

[0027] Determine whether the simulation time reaches the preset value, the flow field residual converges to the set threshold, and the aerodynamic coefficient fluctuation amplitude is lower than the specified range;

[0028] If the simulation time does not reach the preset value, or the flow field residual does not converge to the set threshold, or the aerodynamic coefficient fluctuation amplitude is higher than the specified range, the aerodynamic coefficient calculated by the structure module is fed back to the fluid module to update the flow field boundary conditions, and the fluid-structure coupling calculation is performed again until the simulation time reaches the preset value and the flow field residual and the aerodynamic coefficient meet the convergence requirements;

[0029] Based on the simulation results, the lift coefficient and drag coefficient under different ice thicknesses, different ice types and different wind speed combinations are extracted to generate a correlation data set of ice thickness, wind speed and aerodynamic coefficient.

[0030] Optionally, the calculating of the equivalent aerodynamic coefficient based on the lift coefficient and the drag coefficient, and converting the predicted value of the galloping amplitude of the transmission line under normal operating conditions in combination with a pre-built nonlinear conversion formula includes:

[0031] Perform weighted square fusion calculation based on lift coefficient and drag coefficient to generate equivalent aerodynamic coefficient;

[0032] Input the equivalent aerodynamic coefficient into the pre-built nonlinear conversion formula to obtain the conversion coefficient;

[0033] The predicted value of the galloping amplitude under normal working conditions is dynamically corrected by the conversion coefficient to obtain the predicted value of the galloping amplitude under icing conditions;

[0034] The corrected dancing amplitude prediction value is compared with the preset safety threshold in real time, and a warning signal is triggered when the prediction value exceeds the threshold. At the same time, the prediction value is transmitted to the monitoring center in real time.

[0035] Optionally, in the weighted square fusion calculation, the lift weight factor and the drag weight factor are dynamically adjusted according to the conductor structure parameters;

[0036] The conversion coefficient is calculated by a function containing an exponential parameter, and the exponential parameter is determined by fitting the wind tunnel test data of the ice-covered conductor.

[0037] A second aspect of the present invention provides a system for predicting the galloping amplitude of an ice-covered conductor. Based on the method for predicting the galloping amplitude of an ice-covered conductor described in the first aspect of the present invention, the system comprises:

[0038] Data acquisition module, used to collect environmental meteorological data of transmission lines in real time;

[0039] Data preprocessing module, used to preprocess the data collected in real time;

[0040] Normal operating condition prediction module, used to calculate the predicted value of the galloping amplitude of the transmission line under normal operating conditions;

[0041] Ice thickness prediction module, used to calculate ice thickness prediction value based on pre-processed environmental meteorological data;

[0042] The aerodynamic coefficient calculation module determines the lift coefficient and drag coefficient based on the mapping relationship between ice thickness and aerodynamic coefficients, based on the pre-processed environmental meteorological data and the predicted ice thickness;

[0043] A nonlinear conversion calculation module is used to calculate the equivalent aerodynamic coefficient and convert the predicted value of the transmission line's galloping amplitude under normal operating conditions using a nonlinear conversion formula to obtain the predicted value of the transmission line's galloping amplitude under current icing conditions.

[0044] The communication feedback module is used to transmit the predicted value of the galloping amplitude of the transmission line under the current icing conditions to the monitoring center in real time to achieve remote monitoring and scheduling.

[0045] A third aspect of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for predicting the galloping amplitude of an ice-covered conductor according to the first aspect of the present invention is implemented.

[0046] Compared with the prior art, the beneficial effects of the present invention include at least:

[0047] 1. The present invention adopts a hierarchical modeling strategy, first establishing a baseline model under normal operating conditions, and then achieving precise conversion between ice thickness and aerodynamic characteristics through mapping and nonlinear conversion, thereby significantly improving the accuracy of predictions under icing conditions.

[0048] 2. By combining deep learning with CFD simulation, the system can fully utilize historical and real-time data, and obtain the in-depth impact of icing on the dynamic response of conductors based on physical simulation, thus making up for the lack of prediction accuracy of traditional methods under complex icing conditions.

[0049] 3. The system's integrated design enables efficient and collaborative operation of all modules, providing real-time feedback of forecast results to the dispatching department. This provides a scientific basis for the safe operation of the power grid in icy weather and supports preventive maintenance and emergency control.

[0050] In general, the present invention achieves real-time and accurate prediction of the dancing amplitude of ice-covered conductors through multiple technical means of data-driven, physical simulation and nonlinear conversion, providing an innovative and broadly applicable technical solution for power grid dispatching, fault prevention and system safe operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the overall architecture of the ice-covered conductor galloping amplitude prediction system in an embodiment of the present invention;

[0052] Figure 2 Flowchart of data collection and preprocessing in an embodiment of the present invention;

[0053] Figure 3 This is a CFD simulation flow chart in an embodiment of the present invention;

[0054] Figure 4 This is a flow chart of nonlinear conversion and prediction under icing conditions in an embodiment of the present invention;

[0055] Figure 5 This is an example of the conductor amplitude prediction result in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0057] See Figure 1 A method, system, and storage medium for predicting the galloping amplitude of an ice-covered conductor according to an embodiment of the present invention include the following steps:

[0058] In embodiment 1, the present invention provides a method for predicting the galloping amplitude of an ice-covered conductor, comprising the following steps:

[0059] S1, collecting and preprocessing the environmental meteorological data of the transmission line in real time, wherein the environmental meteorological data includes wind speed, temperature, humidity and precipitation.

[0060] Preferably, if Figure 2 As shown, in S1, the environmental meteorological data of the transmission line includes wind speed W(t), temperature T(t), humidity H(t), precipitation P(t), etc., where t represents time.

[0061] Preferably, in S1, preprocessing the real-time collected data includes:

[0062] Normalization, filtering and denoising, and time synchronization;

[0063] Further preferably, the normalization includes:

[0064] For any parameter data X(t), the statistical normalization formula is used:

[0065]

[0066] where μ X and σ X are the historical mean and standard deviation of the parameter respectively;

[0067] The filtering and denoising include smoothing the data using a sliding average or Kalman filtering method to eliminate instantaneous noise and abnormal fluctuations;

[0068] The time synchronization includes time alignment of data from different sensors to ensure that each data corresponds to the same environmental state at the same time;

[0069] The preprocessed data serves as input for subsequent modules, providing a stable and accurate data foundation for the system.

[0070] S2, inputting the pre-processed environmental meteorological data into a pre-built conductor dynamic response prediction model under normal working conditions to obtain a predicted value of the transmission line's galloping amplitude under normal working conditions.

[0071] Preferably, in S2, the construction of the conductor dynamic response prediction model under normal working conditions includes:

[0072] (1) Building a baseline model

[0073] In the absence of ice, a conductor dynamic response prediction model is established based on historical data.

[0074] Input features: Select wind speed W(t), temperature T(t), humidity H(t) and inherent parameters of the conductor (such as conductor cross-sectional area A c , natural frequency f n ) as model input, exemplarily, the inherent parameters of the conductor include the conductor cross-sectional area A c , natural frequency f n ;

[0075] Output target: Predicted value A of the conductor's galloping amplitude under normal working conditions normal (t).

[0076] (2) Model implementation

[0077] Use fully connected neural network (FCNN) to build a nonlinear mapping model:

[0078] A normal (t) = f θ (W(t),T(t),H(t),Ac(t),f n (t))

[0079] Among them, f θ (·) represents the function determined by the network weight θ;

[0080] The mean square error (MSE) is used as the loss function during training:

[0081]

[0082] where x (i) is the input feature vector of the i-th sample.

[0083] S3, inputting the preprocessed environmental meteorological data into a pre-built ice thickness prediction model to obtain an ice thickness prediction value.

[0084] Preferably, in S3, the construction of the ice thickness prediction model includes:

[0085] Meteorological parameters closely related to conductor icing, such as temperature T(t), humidity H(t), precipitation P(t), and liquid water content L(t), are selected as input features. A deep learning regression model is used to establish a mapping relationship between meteorological parameters and the thickness of ice on the conductor surface I(t):

[0086] I(t)=g φ (T(t),H(t),P(t),L(t)) ω

[0087] where g φ (·) is a regression model with parameter φ, which is trained with a large amount of icing test data, so that the ice thickness of the conductor can be accurately predicted under given current meteorological conditions.

[0088] S4, based on the pre-processed wind speed data and the ice thickness prediction value obtained in S3, the lift coefficient and the drag coefficient are determined according to a pre-established mapping relationship between ice thickness and aerodynamic coefficient.

[0089] Preferably, in S4, establishing a mapping relationship between ice thickness and aerodynamic coefficient includes:

[0090] like Figure 3 As shown, CFD (Computational Fluid Dynamics) software (such as Fluent) is used to perform three-dimensional modeling of the conductor, and aerodynamic simulation is performed under different ice thicknesses I and ice types, including crescent and fan-shaped ice types.

[0091] During the simulation, different wind speed W conditions are set to obtain the corresponding lift coefficient C L (I,W) and drag coefficient C D (I,W).

[0092] Based on the simulation data, the empirical mapping relationship between ice thickness and aerodynamic coefficient is established as follows:

[0093] C L (I,W)=a0+a1I+a2W+a3I 2 +a4W 2 +a5I·W

[0094] C D (I,W)=b0+b1I+b2W+b3I 2 +b4W 2 +b5I·W

[0095] Further preferably, the pre-processed wind speed data and the ice thickness prediction value obtained by S3 are substituted into the established empirical mapping relationship between ice thickness and aerodynamic coefficient to obtain the corresponding lift coefficient C Land the drag coefficient C D .

[0096] Furthermore, Figure 3 The CFD simulation and aerodynamic parameter mapping process shown is one of the core links in the present invention. Through multiple sets of simulations and fitting, the aerodynamic characteristics relationship of the conductor under different ice thickness and wind speed conditions is established.

[0097] Further preferably, the performing aerodynamic simulation by CFD includes:

[0098] System initialization: set parameters such as fluid domain boundary conditions, grid division accuracy and time step;

[0099] Fluid module calculation: Solve the flow field based on the Navier-Stokes equations to obtain the instantaneous aerodynamic data of the ice-covered conductor, including pressure and velocity distribution.

[0100] Import results into the structure module: transfer the aerodynamic data output by the fluid module to the structure module to update the wire dynamics model.

[0101] Aerodynamic data calculation: In the structure module, the material properties of the conductor are combined with the aerodynamic data to calculate the cross-sectional aerodynamic parameters, including lift and drag coefficients.

[0102] Termination condition judgment: If the simulation time reaches the preset termination time, the process ends; otherwise, the calculation results of the structure module are fed back to the fluid module, the flow field boundary conditions are updated, and re-iteration is performed.

[0103] Iterative loop: Achieve high-precision simulation of dynamic aerodynamic characteristics through bidirectional fluid-structure coupling.

[0104] Specifically, the process of performing aerodynamic simulation using CFD includes:

[0105] (1) Conductor geometry modeling

[0106] First, a three-dimensional geometric model is established based on the conductor's dimensions and cross-sectional changes after ice coating.

[0107] Parametric modeling is performed for different ice cover thicknesses and different ice types. The ice cover thicknesses include 0mm, 5mm, and 10mm, and the ice types include crescent and fan.

[0108] (2) Grid division

[0109] The 3D model is meshed using appropriate meshing software, typically unstructured or hybrid meshes. Meshing software such as ICEM and Gambit is used. Mesh refinement is performed on the conductor surface and in critical flow areas to improve the accuracy of capturing key phenomena such as boundary layers and fluid separation.

[0110] (3) Boundary condition setting

[0111] Set the incoming wind speed (W), incoming flow direction and turbulence model (such as k-ω or k-∈), and define the conductor surface as a solid wall with no-slip boundary; set corresponding pressure or symmetry boundary conditions at the outlet, symmetry surface, etc.

[0112] (4) Solution and convergence judgment

[0113] The flow field is solved based on the Navier-Stokes equations to obtain the instantaneous aerodynamic data of the ice-covered conductor, including pressure and velocity distribution. When the flow field changes at the monitoring point tend to be stable and the residual converges to the set threshold, the simulation is considered to have converged. The flow field changes include pressure and velocity.

[0114] (5) Aerodynamic coefficient extraction

[0115] Collect pressure and shear stress distribution on the conductor surface and calculate the lift coefficient C L and the drag coefficient C D ;

[0116] Repeat the above steps to simulate different ice thickness I and wind speed W conditions, and obtain a series of (I, W, C L ,C D ) data sample.

[0117] It should be noted that the existing technology focuses on the static prediction of ice thickness or risk category, and lacks quantitative analysis of the actual dynamic response of the conductor under ice condition (such as dancing amplitude), resulting in a weak correlation between the warning results and the actual vibration state. The present invention establishes a mapping relationship between ice thickness, ice type and aerodynamic coefficient through CFD simulation, combines the dynamic response prediction model to realize quantitative analysis of the dancing amplitude, and realizes the accurate mapping of the aerodynamic characteristics of ice and the dynamic response.

[0118] In this invention, historical meteorological data and historical ice data are used to predict the conductor ice thickness I through a deep learning regression model (such as a fully connected network). The model can automatically learn the complex mapping relationship between meteorological parameters (temperature, humidity, precipitation, etc.) and ice thickness, reducing the limitations of traditional empirical formulas. Based on the predicted ice thickness, the pre-established C L (I,W) and C D The (I, W) mapping relationship can quantify the changes in the aerodynamic performance (lift and drag) of the wire caused by ice cover;

[0119] By combining deep learning with CFD simulation, the team used deep learning to accurately predict ice thickness and CFD results to correct conductor galloping amplitude. This approach ensures the predictions are adaptable to big data while retaining the rigor and accuracy of physical models. The two complement each other and work together to significantly improve the accuracy and practical value of ice-covered conductor galloping amplitude predictions.

[0120] Among them, a i and b i is the parameter obtained by least squares fitting; this mapping relationship provides a physical basis for the subsequent calculation of the conversion coefficient;

[0121] S5 calculates the equivalent aerodynamic coefficient based on the lift coefficient and drag coefficient determined in S4, and then converts the predicted value of the transmission line's dancing amplitude under normal operating conditions using a pre-built nonlinear conversion formula. The predicted value of the transmission line's dancing amplitude under current icing conditions is obtained and transmitted to the monitoring center in real time.

[0122] Preferably, if Figure 4 As shown, the S5 includes:

[0123] In order to convert the baseline prediction results under non-icing conditions into the actual dynamic response under icing conditions, the equivalent aerodynamic coefficient of the conductor under a certain working condition is defined as:

[0124]

[0125] where w L and w D is the weighting factor of lift and drag.

[0126] In the non-icing state, the equivalent aerodynamic coefficient is set to C eq,normal , and in the icing state, the C obtained in S4 is used L (I,W) and C D (I,W) gets C eq,ice .

[0127] The conversion factor γ is defined to convert the dynamic response, and its calculation formula is:

[0128]

[0129] Among them, C eq,normal is the normal working condition reference value of the aerodynamic equivalent coefficient, C eq,ice is the equivalent aerodynamic coefficient generated based on the lift coefficient and drag coefficient calculated by S4, β is an empirical index reflecting the sensitivity of dynamic response, and the relevant parameters are obtained by fitting the experimental data.

[0130] Furthermore, the dynamic response prediction under ice condition is based on the conductor dynamic response A under normal working condition predicted in S2.normal (t), multiplied by the conversion factor γ, the predicted value of the conductor galloping amplitude under icing conditions is obtained:

[0131] A ice (t) = A normal (t)×γ

[0132] It is worth noting that the existing technology does not fully combine the influence of icing on the aerodynamic characteristics of conductors (such as lift and drag) and the dynamic coupling relationship (meteorology, ice layer, structural parameters), and is difficult to deal with the technical problem of nonlinear mutation phenomena in icing evolution (such as rapid thickening or melting of ice layer). The present invention dynamically corrects the normal operating condition prediction value through a nonlinear conversion formula, integrates the aerodynamic parameters (lift and drag coefficient) with real-time meteorological data, and accurately depicts the nonlinear dynamic coupling relationship in icing evolution.

[0133] like Figure 5 The figure shows a comparison between the predicted values ​​(red curve) and the true values ​​(blue curve) of the conductor galloping amplitude over a period of time. The results shown in this figure are based on an example of the measurement phase (or testing phase) of the trained model, which verifies the predictions after given inputs (such as meteorological parameters, conductor parameters, and corresponding conversion factors). In other words, this figure is not a fitting process output in real time during model training, but rather the result obtained when predicting new or partially retained measured data after model training is completed. By comparing the predicted values ​​with the true values, the prediction accuracy and adaptability of the model under actual working conditions can be more intuitively evaluated.

[0134] The following are shown in the figure:

[0135] The overall trend of change is similar: from Figure 5 As can be seen, the predicted curve (red) and the true curve (blue) are relatively consistent in their overall fluctuation patterns, with the fluctuation direction and amplitude trends remaining largely consistent at most time points. This demonstrates that the model can, to a certain extent, reflect the dynamic changes in the conductor's galloping amplitude.

[0136] Local differences still exist: at some time points, there are obvious deviations between the predicted values ​​and the true values. For example, at the peak of the curve or in the area of ​​rapid fluctuation, the predicted curve still has shortcomings in capturing the actual peak or depicting the intensity of fluctuations.

[0137] Room for improvement and optimization: It can be seen that the model has certain predictive capabilities and applicability in practical applications. However, for data with extreme fluctuations in conductor galloping amplitude or complex weather conditions, the conversion formula needs to be further improved, the feature dimension needs to be increased, or the model structure needs to be improved to enhance prediction accuracy and stability.

[0138] In summary, Figure 5A comparison of the predicted and measured results of conductor galloping amplitude during the prediction and verification phase is presented. This figure not only demonstrates that the proposed method can effectively track the changing trends of conductor amplitude in most time periods, but also indicates that there is still room for improvement at certain time points, providing an important basis and reference for subsequent model optimization.

[0139] It is worth noting that the existing technology does not adequately handle the lag effects and dynamic response differences of meteorological parameters during the growth, maintenance and melting stages of ice cover, and does not integrate real-time sensor data with physical simulation results to optimize the risk conversion mechanism, resulting in limited warning accuracy and real-time performance under complex meteorological conditions. The present invention integrates real-time sensor data, CFD simulation results and historical operation data, optimizes the multi-stage correction parameters of ice cover, and uses conversion coefficients to compensate for meteorological errors, significantly improving the robustness of warnings under complex meteorological conditions.

[0140] In embodiment 2, the present invention provides a system for predicting the galloping amplitude of an ice-covered conductor, based on the method for predicting the galloping amplitude of an ice-covered conductor described in embodiment 1. Figure 1 As shown, the system includes:

[0141] Data acquisition module, used to collect environmental meteorological data of transmission lines in real time;

[0142] Data preprocessing module, used to preprocess the real-time collected data, including normalization, filtering and denoising, and time synchronization;

[0143] Normal operating condition prediction module, used to calculate the predicted value of the galloping amplitude of the transmission line under normal operating conditions;

[0144] Ice thickness prediction module, used to calculate ice thickness prediction value based on pre-processed environmental meteorological data;

[0145] The aerodynamic coefficient calculation module determines the lift coefficient and drag coefficient based on the mapping relationship between ice thickness and aerodynamic coefficients, based on the pre-processed environmental meteorological data and the predicted ice thickness;

[0146] A nonlinear conversion calculation module is used to calculate the equivalent aerodynamic coefficient and convert the predicted value of the transmission line's galloping amplitude under normal operating conditions using a nonlinear conversion formula to obtain the predicted value of the transmission line's galloping amplitude under current icing conditions.

[0147] The communication feedback module is used to transmit the predicted value of the galloping amplitude of the transmission line under the current icing conditions to the monitoring center in real time to achieve remote monitoring and scheduling.

[0148] Preferably, each module of the system further includes hardware layout, specifically:

[0149] The data acquisition module includes a meteorological data acquisition unit, which is composed of various sensors. Specifically, meteorological data acquisition sensors are arranged in key sections of the transmission line (such as high-risk icing areas). The meteorological data acquisition unit is equipped with anemometers, thermometers and humidity meters, precipitation sensors, etc., for real-time collection of environmental meteorological parameters.

[0150] The data preprocessing module includes an embedded processor or an edge computing device.

[0151] The normal operating condition prediction module, ice thickness prediction module, aerodynamic coefficient calculation module and nonlinear conversion calculation module include a server or industrial computer platform, running a neural network, a deep regression model and a nonlinear fitting algorithm.

[0152] After system integration, the modules work together to achieve real-time prediction of the dynamic response of conductors under icing conditions, and transmit the prediction results to the dispatching center, providing a scientific basis for power grid accident warning, anti-icing control and maintenance scheduling.

[0153] Embodiment 3 of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for predicting the galloping amplitude of an ice-covered conductor according to embodiment 1 is implemented.

[0154] The present disclosure may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for predicting the galloping amplitude of an ice-covered conductor, characterized in that: The steps include: Real-time collection and pre-processing of environmental meteorological data of transmission lines, including wind speed, temperature, humidity and precipitation; The pre-processed environmental meteorological data is input into the pre-built conductor dynamic response prediction model under normal working conditions to obtain the predicted value of the transmission line's galloping amplitude under normal working conditions; The pre-processed environmental meteorological data is input into the pre-built ice thickness prediction model to obtain the ice thickness prediction value; Based on the pre-processed wind speed data and the predicted ice thickness, the lift coefficient and drag coefficient are determined according to the pre-established mapping relationship between ice thickness and aerodynamic coefficients. The equivalent aerodynamic coefficient is calculated based on the lift coefficient and drag coefficient. The predicted value of the transmission line's dancing amplitude under normal operating conditions is converted using a pre-built nonlinear conversion formula. The predicted value of the transmission line's dancing amplitude under current icing conditions is obtained and transmitted to the monitoring center in real time.

2. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 1, characterized in that: The construction of the conductor dynamic response prediction model under normal working conditions includes: Under non-icing conditions, with wind speed, temperature, humidity and inherent parameters of conductors as input features and conductor dancing amplitude as output target, a dynamic response mapping framework of input features and output targets is established based on historical monitoring data. A fully connected neural network is used to construct a nonlinear mapping model, and the network weight parameters are optimized through the mean square error loss function to achieve functional relationship fitting between input features and dancing amplitude.

3. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 1, characterized in that: The construction of the ice thickness prediction model includes: Temperature, humidity, precipitation, and liquid water content were selected as input features, and the thickness of ice on the conductor surface was used as the output target. A mapping relationship framework between meteorological parameters and ice thickness was established based on the physical mechanism of icing. A deep learning regression model is used to train model parameters through icing test data to achieve nonlinear fitting of the relationship between environmental meteorological conditions and ice thickness.

4. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 1, characterized in that: The establishment of the mapping relationship between ice thickness and aerodynamic coefficient includes: Three-dimensional modeling of the ice-covered conductor was performed using computational fluid dynamics software, and aerodynamic simulations were carried out under different ice thicknesses, ice types, and wind speeds to obtain a data set related to the lift coefficient, drag coefficient, ice thickness, and wind speed. Based on the data set, a six-variable polynomial empirical relationship model of the lift coefficient and drag coefficient was established. Its expression consists of a linear combination of a constant term, a linear / quadratic term of ice thickness, a linear / quadratic term of wind speed, and an ice thickness-wind speed interaction term.

5. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 4, characterized in that: The aerodynamic simulation performed by computational fluid dynamics software includes: Based on the conductor geometric parameters, different ice thicknesses and different ice types, a three-dimensional geometric model of the conductor is constructed and the model is meshed. Set the simulation boundary conditions, including the incoming wind speed, incoming flow direction and turbulence model, and define the conductor surface as a solid wall with no slip; Configure solution parameters, including time step and convergence threshold, and complete simulation environment initialization; In the simulation environment, the Navier-Stokes equations are solved by the fluid module to obtain instantaneous aerodynamic data on the surface of the ice-covered conductor; The instantaneous aerodynamic data output by the fluid module is transferred to the structure module, and the lift coefficient and drag coefficient are calculated based on the material properties of the wire.

6. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 5, characterized in that: The simulation by computational fluid dynamics software also includes: Determine whether the simulation time reaches the preset value, the flow field residual converges to the set threshold, and the aerodynamic coefficient fluctuation amplitude is lower than the specified range; If the simulation time does not reach the preset value, or the flow field residual does not converge to the set threshold, or the aerodynamic coefficient fluctuation amplitude is higher than the specified range, the aerodynamic coefficient calculated by the structure module is fed back to the fluid module to update the flow field boundary conditions, and the fluid-structure coupling calculation is performed again until the simulation time reaches the preset value and the flow field residual and the aerodynamic coefficient meet the convergence requirements; Based on the simulation results, the lift coefficient and drag coefficient under different ice thicknesses, different ice types and different wind speed combinations are extracted to generate a correlation data set of ice thickness, wind speed and aerodynamic coefficient.

7. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 1, characterized in that: The calculation of the equivalent aerodynamic coefficient based on the lift coefficient and the drag coefficient, and the conversion of the predicted value of the galloping amplitude of the transmission line under normal working conditions using a pre-built nonlinear conversion formula include: Perform weighted square fusion calculation based on lift coefficient and drag coefficient to generate equivalent aerodynamic coefficient; Input the equivalent aerodynamic coefficient into the pre-built nonlinear conversion formula to obtain the conversion coefficient; The predicted value of the galloping amplitude under normal working conditions is dynamically corrected by the conversion coefficient to obtain the predicted value of the galloping amplitude under icing conditions; The corrected dancing amplitude prediction value is compared with the preset safety threshold in real time, and a warning signal is triggered when the prediction value exceeds the threshold. At the same time, the prediction value is transmitted to the monitoring center in real time.

8. The method for predicting the galloping amplitude of an ice-covered conductor according to claim 7, characterized in that: In the weighted square fusion calculation, the lift weight factor and the drag weight factor are dynamically adjusted according to the conductor structure parameters; The conversion coefficient is calculated by a function containing an exponential parameter, and the exponential parameter is determined by fitting the wind tunnel test data of the ice-covered conductor.

9. A system for predicting the galloping amplitude of an ice-covered conductor, based on the method for predicting the galloping amplitude of an ice-covered conductor according to any one of claims 1 to 8, characterized in that: The system includes: Data acquisition module, used to collect environmental meteorological data of transmission lines in real time; Data preprocessing module, used to preprocess the data collected in real time; Normal operating condition prediction module, used to calculate the predicted value of the galloping amplitude of the transmission line under normal operating conditions; Ice thickness prediction module, used to calculate ice thickness prediction value based on pre-processed environmental meteorological data; The aerodynamic coefficient calculation module determines the lift coefficient and drag coefficient based on the mapping relationship between ice thickness and aerodynamic coefficients, based on the pre-processed environmental meteorological data and the predicted ice thickness; A nonlinear conversion calculation module is used to calculate the equivalent aerodynamic coefficient and convert the predicted value of the transmission line's galloping amplitude under normal operating conditions using a nonlinear conversion formula to obtain the predicted value of the transmission line's galloping amplitude under current icing conditions. The communication feedback module is used to transmit the predicted value of the galloping amplitude of the transmission line under the current icing conditions to the monitoring center in real time to achieve remote monitoring and scheduling.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, a method for predicting the galloping amplitude of an ice-covered conductor is implemented according to any one of claims 1 to 8.

Images