Numerical simulation of icing on composite insulators and protection and prevention method
By constructing a three-dimensional geometric model of composite insulators, calculating the skirt shading index (SSI), and correcting the simulation boundary conditions, the problem of uneven icing distribution in composite insulators was solved, achieving accurate icing simulation and protection optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHONGQING JIAOTONG UNIV
- Filing Date
- 2025-07-23
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies fail to effectively consider the impact of the shading effect of composite insulator skirts on icing behavior, resulting in inaccurate simulation of uneven icing distribution and a lack of systematic protective measures.
By constructing a three-dimensional geometric model, the shading index SSI of the umbrella skirt is calculated, and the simulation boundary conditions are corrected based on the SSI to simulate the icing growth process of each umbrella skirt layer. The mapping relationship between the degree of shading and the thickness of icing is established by combining a machine learning model to optimize the protective measures.
It enables accurate simulation and protection against non-uniform icing on composite insulators, guides precise protection measures under resource-limited conditions, and improves simulation accuracy and protection efficiency.
Smart Images
Figure CN120850790B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of composite insulator technology, and more specifically, to a numerical simulation and protection method for icing of composite insulators. Background Technology
[0002] In high-voltage transmission lines, composite insulators are widely used to replace traditional porcelain or glass insulators due to their excellent insulation performance, light weight, high strength, and pollution resistance. However, composite insulators in cold or humid regions are highly susceptible to atmospheric icing during operation. Icing not only reduces insulation levels but can also lead to tripping, flashover, and even equipment damage, making it one of the key risk factors for safe power grid operation.
[0003] Composite insulators have a distinctive structural feature: their skirts are typically composed of multiple silicone rubber skirts that are uniformly distributed along the axial direction and are "stacked". This "multi-skirt stacked configuration" enhances insulation against rain, snow and pollution, but it also introduces a new icing mechanism: the "geometric shielding" effect of the upper skirts on the lower skirts.
[0004] Specifically, in natural wind fields, the upper umbrella skirt creates wind speed disturbances and localized low-pressure areas on the lower layer, resulting in a significant area of low wind speed, high humidity, and weak heat transfer on the surface of the lower umbrella skirt. This easily becomes an "anchor point" for the initial attachment of ice crystals. At the same time, because the icing process involves complex gas-liquid-solid phase changes and heat and mass transfer processes, changes in the microenvironment of the shaded area will directly lead to uneven distribution of icing.
[0005] Although existing research has attempted to simulate the insulator icing process using CFD simulations or empirical analysis methods, most current methods do not focus on reflecting the physical mechanism of the "skirt shading effect" during the simulation. In particular, the impact of the skirt structure on icing behavior in multiple dimensions, such as wind field disturbance, droplet trapping, and heat transfer efficiency, lacks systematic modeling and parameterization. Therefore, there is an urgent need for a numerical simulation method that can fully consider the multi-skirt shading effect, accurately reflect the airflow disturbance and heat-mass coupling transfer processes between different skirt levels, and thus achieve simulation of the non-uniform distribution of icing behavior and optimization of protective measures. Summary of the Invention
[0006] The technical problem to be solved by the present invention is to provide a numerical simulation and protection method for icing of composite insulators, so as to solve the problems mentioned in the background art.
[0007] To achieve the above objectives, the present invention adopts the following technical solution:
[0008] The advantages of this invention over the prior art are:
[0009] A numerical simulation and protection method for icing of composite insulators includes the following steps:
[0010] A three-dimensional geometric model of a composite insulator is constructed, the model including multiple awning structures arranged along the axial direction;
[0011] Based on the preset airflow field simulation calculation, the degree of shading corresponding to each umbrella skirt is calculated to obtain the umbrella skirt shading index SSI for each level of umbrella skirt. SSI is a floating-point number with a value between 0 and 1, which is used to represent the strength of the current umbrella skirt being affected by the upper layer of shading.
[0012] Using SSI as an adjustment factor, the simulation boundary condition coefficients are applied to obtain the corrected simulation boundary condition coefficients for each umbrella skirt surface.
[0013] Simulations were performed based on the corrected simulation boundary condition coefficients to simulate the icing growth process of each level of the umbrella skirt.
[0014] Output the icing thickness distribution results at the umbrella skirt level to analyze the influence of the shading effect on icing non-uniformity.
[0015] Preferably, the process for calculating SSI is as follows:
[0016] Fluid dynamics simulation was performed on a three-dimensional model of the composite insulator under ice-free conditions to obtain the actual local wind speed v on the surface of each skirt. n ;
[0017] And define the umbrella skirt occlusion index SSI n for:
[0018] SSI n = 1 - (v n / v nr );
[0019] Among them, v nr The reference wind speed for the nth layer refers to the reference wind speed of the nth layer umbrella skirt under unobstructed conditions.
[0020] Preferably, the calculation steps for the reference wind speed of the nth layer of the umbrella skirt include:
[0021] While keeping the original simulation boundary conditions unchanged, a simplified geometric model of the composite insulator is constructed, in which all umbrella skirt structures located above the nth layer of umbrella skirts are removed from the simplified model;
[0022] Based on the simplified geometric model, a fluid dynamics simulation under ice-free conditions was performed to obtain the wind speed on the surface region of the nth layer of the umbrella skirt as the reference wind speed v. nr .
[0023] Preferably, the formula for applying SSI as an adjustment factor to the simulation boundary condition coefficients includes:
[0024] h n = h nr × (1 - α × SSI n );
[0025] h nr This represents the simulation boundary condition coefficients of the nth layer of the umbrella skirt in the unobstructed state;
[0026] h n This represents the simulation boundary condition coefficients after the nth layer of the umbrella skirt is corrected.
[0027] Preferably, the simulation boundary condition coefficients include any one or more of the following: surface heat transfer coefficient, wind speed, droplet capture efficiency, and evaporation rate.
[0028] Preferably, the simulation process is implemented using simulation software, including any one or more of the following: Fluent, OpenFOAM, COMSOL Multiphysics, and ANSYS CFX.
[0029] Preferably, the steps for analyzing the effect of shading on icing nonuniformity include:
[0030] Establish a mapping relationship between each umbrella skirt number and its corresponding simulated icing thickness data to generate an umbrella skirt level-icing thickness distribution table;
[0031] Based on the umbrella skirt layer-icing thickness distribution table, the shading index SSI is constructed. n With ice thickness H n The fitting curve or correlation coefficient between the two is used to quantitatively assess the impact of the degree of shading on the degree of icing.
[0032] Preferably, the method further includes: outputting an ice thickness distribution map or an occlusion-icing effect map of the umbrella skirt layer through visualization means.
[0033] Preferably, the method further includes: constructing a training sample set based on the output distribution results of icing thickness of each umbrella skirt, combined with the corresponding umbrella skirt occlusion index (SSI), and using a supervised learning model to model the mapping relationship between the degree of occlusion and the non-uniformity of icing.
[0034] Preferably, the method further includes: prioritizing the protection and prevention of the umbrella skirt based on the icing thickness.
[0035] This invention fully considers the geometric structural characteristics of composite insulators with multiple skirts stacked along the axial direction. For the first time, it introduces the parameter "skirt shading index (SSI)" into the simulation model to quantitatively characterize the wind speed attenuation and heat exchange interference caused by the upper skirt to the lower skirt, thus overcoming the limitations of traditional simulations in simplifying the multi-level skirt structure and inaccurate prediction of icing distribution.
[0036] This invention obtains the wind speed values of each umbrella skirt affected by the flow by performing fluid dynamics simulation in an ice-free state, and accurately calculates the SSI value of each umbrella skirt by combining the reference wind speed under unobstructed conditions. This allows the simulation to apply differentiated boundary conditions to different umbrella skirts, such as surface heat transfer coefficient, evaporation rate, droplet capture efficiency, etc., thereby more realistically reproducing the local non-uniformity characteristics of the icing process.
[0037] This invention utilizes modified simulation boundary conditions to simulate icing growth, which can not only output the icing thickness distribution results of each umbrella skirt layer, but also further analyze the influence of the shading effect on the non-uniformity of icing, thus making up for the shortcomings of existing models in terms of spatial resolution and local mechanism fitting ability.
[0038] This invention also supports combining simulation output results with machine learning models to construct a nonlinear mapping relationship between the shading index and icing thickness, which is used to establish a predictive model of structure-environment-icing behavior. This expands the data-driven capability of traditional CFD simulation and helps to form a more intelligent icing analysis framework.
[0039] This invention can further automatically generate a priority ranking of umbrella skirt layer protection based on simulated icing thickness results, thereby guiding the deployment of measures such as heating, coating, and structural fine-tuning, and achieving precise protection under conditions of limited resources. Attached Figure Description
[0040] Figure 1 This is the overall flowchart of the method of the present invention;
[0041] Figure 2 This is a flowchart of the method for calculating SSI according to the present invention. Detailed Implementation
[0042] The specific embodiments of the present invention will now be described with reference to the accompanying drawings.
[0043] like Figure 1 As shown, the method of the present invention specifically includes the following steps:
[0044] A three-dimensional geometric model of a composite insulator is constructed, the model including multiple awning structures arranged along the axial direction;
[0045] Based on the preset airflow field simulation calculation, the degree of shading corresponding to each umbrella skirt is calculated to obtain the Skirt Shading Index (SSI) of each umbrella skirt. SSI is a floating-point number with a value between 0 and 1, which is used to represent the strength of the current umbrella skirt being affected by the upper layer of shading.
[0046] Using SSI as an adjustment factor, the simulation boundary condition coefficients are applied to obtain the corrected simulation boundary condition coefficients for each umbrella skirt surface.
[0047] Simulations were performed based on the corrected simulation boundary condition coefficients to simulate the icing growth process of each level of the umbrella skirt.
[0048] Output the icing thickness distribution results at the umbrella skirt level to analyze the influence of the shading effect on icing non-uniformity.
[0049] Composite insulators consist of a core rod and multiple sheds arranged axially. The geometry and arrangement of the sheds significantly affect the airflow field and icing process. Therefore, the model needs to accurately reflect the structural characteristics of actual insulators.
[0050] In practice, computer-aided design software (such as SolidWorks, CATIA) or 3D modeling tools can be used to create the model. During modeling, parameters such as the diameter, thickness, and spacing of the skirts must be input based on design drawings or physical measurement data. For example, suppose a composite insulator contains 10 skirts with diameters ranging from 100mm, 110mm, 120mm to 190mm from top to bottom, a spacing of 50mm, and a core rod diameter of 30mm. During modeling, minor details (such as small chamfers) can be ignored, but the overall shape and relative positions of the skirts must be preserved to ensure the accuracy of subsequent airflow field simulations. After the model is completed, it needs to be exported to a format suitable for simulation software, such as STL or IGES, for import into computational fluid dynamics (CFD) tools for subsequent analysis.
[0051] The Skirt Shading Index (SSI) is an indicator used to quantify the degree to which the current skirt is affected by the shading of the upper skirt, and its value ranges from 0 to 1.
[0052] like Figure 2 As shown, the SSI calculation is based on fluid dynamics simulation under ice-free conditions, and the shading effect is determined by comparing the local wind speed with the reference wind speed.
[0053] First, a fluid dynamics simulation under ice-free conditions is performed on the complete 3D model to obtain the local actual wind speed v on each umbrella skirt surface. nSimulations can be performed using CFD software such as Fluent, OpenFOAM, COMSOL Multiphysics, or ANSYS CFX. When setting up the simulation, boundary conditions are defined, including setting the inlet wind speed (e.g., 5 m / s), and the turbulence model can be a k-ε model, etc. After the simulation is complete, the average wind speed on the surface of each skirt layer is extracted as v. n For example, for the third layer of the umbrella skirt, assume that its surface average wind speed v3 is 4.2 m / s.
[0054] Next, calculate the reference wind speed v. nr This represents the wind speed of the nth layer of the umbrella skirt without any upper layer obstructing it, where the subscript 'r' indicates a reference value. To address this, a simplified model is constructed by removing all umbrella skirt structures above the nth layer. For example, if calculating the v of the 3rd layer umbrella skirt... 3r If the first and second layers of the umbrella skirt are removed, and layers 3 through 10 are retained, then a fluid dynamics simulation is performed on the simplified model under the same boundary conditions as the original simulation. The average wind speed on the surface of the third layer of the umbrella skirt is extracted, and the result is assumed to be v. 3r =5m / s.
[0055] Based on local wind speed v n and reference wind speed v nr Umbrella skirt coverage index (SSI) n Defined as:
[0056] ;
[0057] Taking the third-layer umbrella skirt as an example, if v3 = 4.2 m / s, v 3r =5m / s, then:
[0058] ;
[0059] This formula is designed based on the following logic: when the local wind speed is v n Approximate reference wind speed v nr When the upper layer shading has a relatively small impact, it indicates that the SSI is relatively stable. n Approaching 0; when v n Much smaller than v nr When this occurs, it indicates a significant blocking effect, weakening the airflow and reducing SSI. n Approaching 1.
[0060] This definition method can intuitively reflect the impact of shielding on the airflow field and provide a quantitative basis for subsequent boundary condition correction.
[0061] In icing simulations, boundary condition coefficients (such as surface heat transfer coefficient, wind speed, droplet trapping efficiency, and evaporation rate) directly affect the icing growth process. Since shading effects can alter local conditions, these coefficients need to be corrected based on the SSI (Surface Temperature Index).
[0062] The corrected formula is:
[0063] ;
[0064] in:
[0065] These are the boundary condition coefficients after correction for the nth layer of the umbrella skirt;
[0066] The boundary condition coefficients for the nth layer of the umbrella skirt in the unobstructed state;
[0067] This is the adjustment coefficient.
[0068] The significance of this formula lies in adjusting the boundary condition coefficients according to the degree of occlusion. When SSI... n A larger value indicates severe shading, reduced local wind speed, and weakened heat exchange capacity or droplet capture efficiency. It should be reduced; when SSI n When the size is small, the impact of shading is minimal. near The multiplicative form ensures that the corrected coefficients maintain a proportional relationship with the original values, while This controls the degree of influence of the shading effect.
[0069] These are boundary condition coefficients under unobstructed conditions, which can be obtained through theoretical calculations or simulations of a single-layer umbrella skirt. For example, the surface heat transfer coefficient can be estimated based on the convection heat transfer formula or directly extracted through CFD simulation.
[0070] The value is generally between 0 and 1, with the specific value depending on the actual impact of the occlusion effect on the boundary conditions. It can be calibrated using experimental data. If there is no experimental support, it can be initially set to 0.5 and adjusted according to the accuracy of the simulation results. The more accurate the simulation results, the better the parameter selection.
[0071] Based on the corrected boundary condition coefficients, numerical simulations of the ice growth process can be further performed. The ice formation process involves physical phenomena such as droplet impact, freezing, and heat conduction, requiring the use of software that supports ice formation simulation, such as Fluent's FENSAP-ICE module or a custom solver for OpenFOAM.
[0072] The following parameters need to be set for simulation:
[0073] Environmental conditions: temperature (e.g., -5℃), droplet diameter (e.g., 20μm), droplet concentration (e.g., 1g / m³).
[0074] Boundary conditions: the modified Applied to the surface of various umbrella skirts.
[0075] The physical model includes droplet motion (using Eulerian or Lagrangian methods), freezing process (considering phase change heat), heat conduction, etc.
[0076] For example, in Fluent, this can be achieved through user-defined functions (UDFs). The system dynamically adjusts and runs an icing simulation for a certain period of time (e.g., 30 minutes). The simulation process of icing is existing technology, so it will not be described in detail.
[0077] After the simulation is completed, the icing thickness distribution of each umbrella skirt is output. For example, the thickness of the first umbrella skirt is H1=2mm, the thickness of the second skirt is H2=2.5mm, and so on.
[0078] Based on simulation results, the impact of shading effect on icing distribution is analyzed. First, a skirt-level icing thickness distribution table is generated. Then, an SSI (Surface Area Measurement System) is constructed. n With H n The fitting relationship between them. Using Python's SciPy library, a linear model can be fitted:
[0079] ;
[0080] and These are undetermined coefficients. SSI is plotted using Matplotlib. n With H n The scatter plot and fitted curve can visually demonstrate the impact of the occlusion effect.
[0081] To further investigate the relationship between occlusion and icing, a training sample set can be constructed, and a supervised learning model can be used for prediction. Using SSI... n and H n For example, the sample set may contain 20 sets of data.
[0082] We choose a simple neural network (MLP) with the following architecture:
[0083] The input layer consists of one neuron (input SSI). n );
[0084] The hidden layer consists of two layers, each with 10 neurons, using the ReLU activation function;
[0085] The output layer consists of one neuron (output H). n( ), with no activation function.
[0086] Training the model using Python's TensorFlow library:
[0087] The loss function is the mean squared error (MSE).
[0088] The optimizer is Adam;
[0089] The number of training rounds is set to 1000.
[0090] The dataset is divided into 80% for training and 20% for testing.
[0091] After training, the model can predict a given SSI. n H n For example, if the input SSI=0.2, the predicted H=2.8mm.
[0092] The umbrella skirts are ordered according to the icing thickness. For example, if H5=4mm, H4=3.5mm, and H1=2mm, then the priority is umbrella skirt 5> umbrella skirt 4>...> umbrella skirt 1.
[0093] For high-priority umbrella skirts, measures such as heating or anti-icing coatings can be prioritized.
[0094] This invention fully considers the geometric structural characteristics of composite insulators with multiple sheds stacked along the axial direction, and for the first time introduces the shed shading index parameter into the simulation model, thereby improving the accuracy and efficiency of the simulation.
[0095] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A numerical simulation and protection method for icing of composite insulators, characterized in that, Includes the following steps: A three-dimensional geometric model of a composite insulator is constructed, the model including multiple awning structures arranged along the axial direction; Based on the preset airflow field simulation calculation, the degree of shading corresponding to each umbrella skirt is calculated to obtain the umbrella skirt shading index SSI for each level of umbrella skirt. SSI is a floating-point number with a value between 0 and 1, which is used to represent the strength of the current umbrella skirt being affected by the upper layer of shading. Using SSI as an adjustment factor, the simulation boundary condition coefficients are applied to obtain the corrected simulation boundary condition coefficients for each umbrella skirt surface. Simulations were performed based on the corrected simulation boundary condition coefficients to simulate the icing growth process of each level of the umbrella skirt. Output the icing thickness distribution results at the umbrella skirt level to analyze the influence of the shading effect on icing non-uniformity; The process of calculating SSI is as follows: Fluid dynamics simulation was performed on a three-dimensional model of the composite insulator under ice-free conditions to obtain the actual local wind speed v on the surface of each skirt. n ; And define the umbrella skirt occlusion index SSI n for: SSI n =1-(v n / v nr ); Among them, v nr The reference wind speed for the nth layer refers to the reference wind speed of the nth layer umbrella skirt under unobstructed conditions. The steps for calculating the reference wind speed for the nth layer of the umbrella skirt include: While keeping the original simulation boundary conditions unchanged, a simplified geometric model of the composite insulator is constructed, in which all umbrella skirt structures located above the nth layer of umbrella skirts are removed from the simplified geometric model. Based on the simplified geometric model, a fluid dynamics simulation under ice-free conditions is performed to obtain the wind speed on the surface region of the nth layer of the umbrella skirt as the reference wind speed v. nr .
2. The numerical simulation and protection method for icing of composite insulators according to claim 1, characterized in that, The formulas for SSI as an adjustment factor affecting the simulation boundary condition coefficients include: h n =h nr ×(1-α×SSI n ); h nr This represents the simulation boundary condition coefficients of the nth layer of the umbrella skirt in the unobstructed state; h n α represents the simulation boundary condition coefficients after the nth layer of the umbrella skirt is corrected, and α is the adjustment coefficient.
3. The numerical simulation and protection method for icing of composite insulators according to claim 2, characterized in that, The simulation boundary condition coefficients include any one or more of the following: surface heat transfer coefficient, wind speed, droplet capture efficiency, and evaporation rate.
4. The numerical simulation and protection method for icing of composite insulators according to claim 1, characterized in that, The simulation process is implemented using simulation software, including any one or more of the following: Fluent, OpenFOAM, COMSOL Multiphysics, and ANSYS CFX.
5. The numerical simulation and protection method for icing of composite insulators according to claim 1, characterized in that, The steps for analyzing the impact of shading effects on icing nonuniformity include: Establish a mapping relationship between each umbrella skirt number and its corresponding simulated icing thickness data to generate an umbrella skirt level-icing thickness distribution table; Based on the umbrella skirt layer-icing thickness distribution table, the shading index SSI is constructed. n With ice thickness H n The fitting curve or correlation coefficient between the two is used to quantitatively assess the impact of the degree of shading on the degree of icing.
6. The numerical simulation and protection method for icing of composite insulators according to claim 5, characterized in that, The method also includes: outputting a map of ice thickness distribution at the umbrella skirt level or a map of the effects of shading and icing through visualization.
7. The numerical simulation and protection method for icing of composite insulators according to claim 5, characterized in that, The method further includes: constructing a training sample set based on the output distribution results of icing thickness of each umbrella skirt, combined with the corresponding umbrella skirt occlusion index (SSI), and using a supervised learning model to model the mapping relationship between the degree of occlusion and the non-uniformity of icing.
8. The numerical simulation and protection method for icing of composite insulators according to claim 5 or 7, characterized in that, The method further includes: prioritizing the protection and prevention of the umbrella skirt based on the icing thickness.