Power transmission line icing defect physical simulation sample generation method and related device

Abstract

The invention belongs to a sample generation method, and provides a physical simulation sample generation method for power transmission line icing defects and a related device for solving the technical problems that an existing method for generating power transmission line icing samples in a simulation mode is insufficient in authenticity, diversity and high efficiency. Calculating an icing defect generation path of the power transmission line, performing parameterization generation on an ice column model, determining an attachment position of an ice column generated by parameterization on the power transmission line in combination with a probability density field and inverse transformation sampling, enabling the ice column to present a windward deflection form, and enabling the ice column to be attached to the icing defect generation path of the power transmission line, the method obtains the icing defect virtual scene, converts the icing defect virtual scene into available sample data, and is of great significance in improving the icing defect early warning capability of a power system and guaranteeing safe and stable operation of a power grid.

Description

Technical Field

[0001] This invention pertains to a sample generation method, specifically a method and apparatus for generating physical simulation samples of icing defects in power transmission lines. Background Technology

[0002] Icing on power transmission lines is a common and serious hidden danger in power systems during winter, easily leading to accidents such as line flashover, line breaks, and tower collapses, directly threatening the safe and stable operation of the power grid. In order to provide early warning of icing defects and optimize de-icing maintenance strategies, it is necessary to train detection models based on a large amount of icing sample data, and the authenticity and diversity of the samples directly determine the detection accuracy.

[0003] Existing methods for obtaining icing samples mainly fall into two categories: field collection and simulation generation. Field collection requires natural icing conditions, is highly limited by region, climate, and season, has a long cycle, high cost, and struggles to cover extreme scenarios with different wind speeds, lighting, and icing patterns, resulting in a severe lack of sample quantity and diversity. Traditional simulation generation methods are mostly based on simple geometric modeling, which can only simulate regular icing patterns and cannot reproduce the influence of wind direction on the attachment position and deflection pattern of icing in the real environment. Furthermore, they lack physical constraints, and the icing distribution does not match the actual sag path and curvature characteristics of the transmission line, resulting in poor sample realism and difficulty in meeting the training requirements of detection models.

[0004] Furthermore, existing simulation schemes lack a scientific mechanism for the probability distribution of icing, resulting in highly random icing attachment locations that are disconnected from actual icing growth patterns. Additionally, the flexibility in adjusting environmental parameters is low, making it impossible to generate high-fidelity icing samples under multiple operating conditions in batches. Therefore, there is an urgent need for a physical simulation sample generation method for power transmission line icing defects that can balance realism, diversity, and efficiency, thereby overcoming the current technological bottlenecks. Summary of the Invention

[0005] This invention addresses the technical problems of insufficient realism, diversity, and efficiency in existing methods for simulating and generating icing samples of power transmission lines. It provides a method and related apparatus for generating physical simulation samples of icing defects in power transmission lines.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention proposes a method for generating physical simulation samples of icing defects in power transmission lines, comprising: Construct models of electrical equipment and various forms of icicles respectively; The power equipment model and icicle model were imported into Unreal Engine to obtain a virtual scene of the power transmission line; Set the environmental parameters for the virtual scene of the power transmission line; Based on the power equipment model, icicle model and environmental parameters, the generation path of icing defects on the transmission line is calculated, and the icicle model is parameterized. Combining the probability density field and inverse transformation sampling, the attachment position of the parameterized icicle on the transmission line is determined, and the icicle is made to present a windward deflection shape. By attaching icicles to the path of icing defects in the power transmission line, a virtual scene of icing defects is obtained. The virtual scenario of icing defects is transformed into usable sample data.

[0007] Furthermore, methods for constructing power equipment models include: Point cloud data and image data of the power transmission line were collected using a 3D scanner and a visible light camera. Three-dimensional models of transmission towers, transmission lines, and connecting equipment are reconstructed based on point cloud data and image data.

[0008] Furthermore, the method for importing the power equipment model and the icicle model into Unreal Engine to obtain a virtual scene of the power transmission line includes: Import the power equipment model and the icicle model into Unreal Engine, add materials to the power equipment model and the icicle model, and then set the collision volume for the power equipment model and the icicle model.

[0009] Furthermore, the method for setting the environmental parameters of the virtual scene of the transmission line includes: Set the weather conditions in the environment to snow and cover it with snow; Simulate lighting conditions at any time of day using the lighting system in Unreal Engine; Set environmental data for wind speed and wind direction.

[0010] Furthermore, the method for calculating the generation path of icing defects in the transmission line includes: Estimate the maximum sag at the midpoint of the transmission line :

[0011] in, This refers to the weight per unit length of the power transmission line. The horizontal tension of the transmission line. This refers to the horizontal span between the two suspension ends of the power transmission line. Starting point of the target transmission line suspension endpoint Establish a local coordinate system with the origin and sample at equal intervals along the horizontal span direction. For each sampling point, calculate the height corresponding to the local coordinates. :

[0012] in, For the first The x-coordinate of the local coordinates of each sampling point; The vertical height difference between the two suspension ends of the target transmission line; All calculated path points are transformed from the local coordinate system to the world coordinate system of the 3D scene to obtain a 3D point set; the x-coordinate of the path point is the x-coordinate of the local coordinate of the sampling point, and the y-coordinate is the corresponding height. Based on the 3D point set, a 3D spline curve is generated using the spline curve component in Unreal Engine, which serves as the path for generating icing defects.

[0013] Furthermore, the method for determining the attachment location of the parameterized icicles on the power transmission line by combining probability density field and inverse transform sampling includes: The path of icing defect generation is sampled, with arc length as the data. Calculate the sampling points for each parameter. curvature ; Calculate the ice accretion probability density field :

[0014] in, It is the angle between the wind direction and the normal direction of the power transmission line. and All are weighting coefficients. For global wind speed; For the ice accretion probability density field Normalization is performed to obtain the standard probability density function; By combining the standard probability density function, the cumulative distribution function of the icing probability density field is calculated; Based on the snow level in the virtual scene, determine the total number of icicles to be generated and generate uniformly distributed random numbers accordingly. For each random number, sample it by inverse transformation of the cumulative distribution function of the ice accretion probability density field and calculate the corresponding position of the random number on the ice accretion defect generation path.

[0015] Furthermore, the method for causing the icicle to deflect in the wind includes: Determine the deflection axis and deflection angle of the icicle respectively; By changing the angle of rotation around the deflection axis, the icicle is made to deflect in the wind.

[0016] Secondly, this invention proposes a physical simulation sample generation system for transmission line icing defects, comprising: The model module is used to build power equipment models and various types of icicle models, respectively. The virtual module is used to import power equipment models and icicle models into Unreal Engine to obtain a virtual scene of power transmission lines; The parameter module is used to set the environmental parameters of the virtual scene of the power transmission line; The calculation module is used to calculate the icing defect generation path of the transmission line based on the power equipment model, icicle model and environmental parameters, and to parameterize the icicle model. Combining the probability density field and inverse transformation sampling, it determines the attachment position of the parameterized icicle on the transmission line and makes the icicle present a windward deflection shape. Combined with the module, it is used to make icicles adhere to the path of icing defects in power transmission lines, thereby obtaining a virtual scene of icing defects. The conversion module is used to convert virtual scenes of icing defects into usable sample data.

[0017] Thirdly, the present invention proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-mentioned method for generating physical simulation samples of icing defects in transmission lines.

[0018] Fourthly, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for generating physical simulation samples of icing defects in transmission lines.

[0019] Compared with the prior art, the present invention has the following beneficial technical effects: This invention proposes a physical simulation method for generating samples of icing defects on power transmission lines. Based on power equipment models, icicle models, and environmental parameters, the method calculates the icing defect generation path of the power transmission line and parametrically generates the icicle model. Combining probability density fields and inverse transformation sampling, the method determines the attachment position of the parametrically generated icicles on the power transmission line and makes the icicles exhibit a wind-facing deflection shape. The icicles are then attached to the icing defect generation path of the power transmission line to obtain a virtual scene of icing defects, which is then converted into usable sample data. This invention does not rely on natural icing conditions or on-site collection. Through virtual simulation, it can generate icing samples in batches under multiple operating conditions, environments, and morphologies. Combined with automatic annotation, it significantly shortens the sample acquisition cycle, reduces collection costs, and solves the problem of difficulty in capturing samples under extreme operating conditions. By adjusting model parameters, environmental parameters, and simulation algorithms, it can adapt to the needs of generating icing samples for power transmission lines under different voltage levels, terrains, and climatic conditions, making it widely applicable. It can be used not only for training icing defect detection models, but also for generating simulation samples of other defects in transmission lines. It has strong technical versatility and promotional value, and is of great significance for improving the early warning capability of power system icing defects and ensuring the safe and stable operation of the power grid.

[0020] This invention also proposes a physical simulation sample generation system for transmission line icing defects, an electronic device, and a computer-readable storage medium, which possesses all the advantages of the aforementioned physical simulation sample generation method for transmission line icing defects. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a method for generating physical simulation samples of icing defects in power transmission lines according to the present invention. Figure 2 This is another flowchart illustrating the method for generating physical simulation samples of icing defects in power transmission lines according to the present invention. Figure 3 This is a schematic diagram illustrating a process for generating icing defects in power transmission lines according to an embodiment of the present invention. Figure 4 This is a schematic diagram of a physical simulation sample generation system for icing defects in power transmission lines according to the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0023] Icing defect detection on power transmission lines is a core technology in the field of power system operation and maintenance, widely used in winter safety protection of high-voltage and ultra-high-voltage transmission lines. As a fundamental energy infrastructure, power transmission lines traverse vast areas, often passing through complex terrains such as mountains and hills. Low temperatures, rain, snow, and strong winds in winter easily lead to icing on the surface of these lines. Icing alters the mechanical and electrical properties of transmission lines, affecting not only normal power transmission but also being a major cause of grid faults. It directly impacts the safe and stable operation of the power system and the guarantee of electricity supply for society. Therefore, accurate early warning and efficient handling of icing defects on power transmission lines have always been a key research topic in the field of power operation and maintenance. Currently, computer vision-based icing defect detection technology is widely used due to its high efficiency and real-time advantages. Its core is to identify icing defects by training a detection model. The quality of model training depends entirely on a large amount of icing sample data. The samples must possess both high realism and rich diversity to ensure reliable recognition accuracy of the detection model under complex actual operating conditions.

[0024] In practical power operation and maintenance, the core challenge in detecting icing defects on transmission lines is acquiring high-quality sample data. On one hand, icing is highly seasonal and geographically specific, occurring only during winter's low temperatures and snow. Furthermore, influenced by topography and climate, the morphology, thickness, and distribution of icing vary significantly across different regions. On the other hand, extreme icing conditions (such as irregular icing accompanied by strong winds and blizzards) have a low probability of occurrence and are difficult to capture on-site, resulting in a severe shortage of diverse samples needed for training detection models. Simultaneously, the development of icing defects is dynamic; the line conditions and environmental parameters differ at different icing stages. If the samples cannot cover the icing characteristics throughout the entire lifecycle, it will directly affect the detection model's ability to identify early-stage icing defects, thus delaying de-icing maintenance and increasing the risk of line failures.

[0025] To address the challenge of obtaining icing samples, existing technologies primarily employ two solutions: on-site collection and simulation generation. On-site collection involves deploying monitoring equipment in high-icing areas or manually taking photos on-site to directly acquire data on real icing scenes. This method maximizes sample authenticity. Simulation generation, on the other hand, utilizes 3D modeling and simulation technology to construct power line models and icing scenes in a virtual environment, generating sample data in batches. Commonly used modeling tools include 3ds Max, while simulation engines often employ platforms with physical rendering capabilities to enhance the visual realism of the samples. Each approach has its focus: on-site collection ensures sample fidelity, while simulation generation attempts to overcome the limitations of natural conditions and compensate for insufficient sample quantity and diversity.

[0026] However, the two existing solutions still have significant shortcomings and cannot meet the training requirements of high-precision detection models. Field data collection is heavily constrained by location, climate, and season, making it not only time-consuming, labor-intensive, and costly, but also unable to actively acquire samples from extreme icing conditions, resulting in limited sample coverage and difficulty in forming a large-scale sample library. Traditional simulation generation methods have many technical shortcomings. They are mostly based on simple geometric modeling, only able to simulate regular icing patterns, and cannot reproduce the influence of wind direction on the attachment position and deflection shape of icing in the real environment. Furthermore, they lack physical constraints, failing to consider the sag path of transmission lines due to their own weight and the curvature differences at different locations, leading to a mismatch between icing distribution and actual conditions, and poor sample realism. In addition, existing simulation schemes have not established a scientific probability distribution mechanism that fits the actual icing growth pattern; the icicle attachment position is highly random, and the flexibility of adjusting scene environmental parameters (such as illumination, wind speed, and snow thickness) is low, making it impossible to generate high-fidelity icing samples under multiple conditions in batches. It is difficult to balance sample realism, diversity, and generation efficiency; this technical bottleneck urgently needs to be overcome.

[0027] Based on the above, this invention proposes a method and related apparatus for generating physical simulation samples of icing defects in power transmission lines. The invention will be further described in detail below with reference to embodiments and accompanying drawings.

[0028] like Figure 1 The diagram shown is a flowchart illustrating a method for generating physical simulation samples of icing defects in power transmission lines according to the present invention, which may include: S101, construct power equipment models and various icicle models respectively.

[0029] It should be noted that the power equipment model specifically refers to the 3D digital model of core power components such as transmission towers, transmission lines, and connecting fittings, while the icicle model is a 3D geometric model simulating the natural icing morphology. It is essential to ensure the geometric consistency between the power equipment model and the actual object, while also using various icicle models to cover different icing growth characteristics. In practical applications, a combination of 3D scanning and image acquisition can be used for modeling. This involves acquiring point cloud data using a 3D scanner and texture data using a visible light camera, then reconstructing the 3D structure of the power equipment in modeling software such as 3ds Max and Maya. Simultaneously, combined with measured data from natural icing, icicle models of different thicknesses, lengths, and curvatures are designed to build a basic model library. This provides a high-fidelity core carrier for subsequent virtual scene construction, avoiding simulation deviations caused by model distortion.

[0030] S102, import the power equipment model and icicle model into Unreal Engine to obtain a virtual scene of the power transmission line.

[0031] It's important to note that Unreal Engine is a real-time 3D creation tool with high-fidelity rendering and physical simulation capabilities. The power transmission line virtual scene refers to a digital virtual space integrating power equipment, icicle models, and the basic environment. Utilizing Unreal Engine's model import and scene integration capabilities, models generated by external modeling software are transformed into editable and simulateable scene elements. Simultaneously, physical properties and visual textures are given to the models, constructing a virtual environment with basic interactivity. In practical applications, the power equipment and icicle models can first be imported into Unreal Engine. Materials are then added to the models using a physically based rendering material system. The power equipment materials can be created based on real-world texture maps, while the icicle materials simulate the texture of real ice by adjusting refractive index and roughness parameters. Collision volumes are then set for each model to ensure that subsequent physical interactions function correctly, ultimately integrating them to form the power transmission line virtual scene.

[0032] S103, set the environmental parameters of the virtual scene of the power transmission line.

[0033] This invention simulates the real-world environmental conditions of natural icing, replicating different operating conditions by adjusting environmental parameters, thus enhancing sample diversity and simulation realism. This is because the morphology and distribution of icing are directly affected by environmental factors such as weather, lighting, and wind. In practical applications, the Unreal Engine's built-in weather system can be used to set the scene to snow and cover it with snow, recreating the basic weather background for icing. The engine's lighting system allows adjustment of the light source angle, intensity, and color temperature to simulate lighting conditions at any time of day, enriching the sample lighting scenarios. Furthermore, global wind speed and direction parameters can be manually set, allowing for different wind speed levels and angles to be set according to actual icing conditions, providing data support for subsequent simulations of icicle deflection. Therefore, this invention constructs an environmental background consistent with real icing conditions, making subsequent icing simulations more realistic. Simultaneously, multi-parameter control achieves sample environmental diversity, avoiding the limitations of a single environmental sample.

[0034] S104. Based on the power equipment model, icicle model and environmental parameters, calculate the icing defect generation path of the transmission line, and parameterize the icicle model. Combine the probability density field and inverse transformation sampling to determine the attachment position of the parameterized icicle on the transmission line and make the icicle present a windward deflection shape.

[0035] This invention integrates mechanical calculations and probabilistic statistical methods to achieve accurate and realistic simulation of icing defects. It ensures that the icicle attachment path conforms to the physical shape of the power transmission line, and uses probability distribution to ensure that the icing location follows natural growth patterns. Simultaneously, it incorporates wind parameters to simulate the stress pattern of the icicle. In practical applications, the icing defect generation path is first calculated. Based on physical parameters such as the weight per unit length of the power transmission line and horizontal tension, a parabolic approximation method is used to estimate the natural sag curve of the power transmission line. A local coordinate system is established with the suspension end of the power transmission line as the origin. Path points are sampled at equal intervals and transformed to the world coordinate system. A smooth three-dimensional path is generated using the Unreal Engine spline curve component as the icing attachment benchmark. Next, icicles are generated parametrically. Models are randomly selected from an icicle model library, and the scaling, rotation, and transparency are randomly adjusted within a fixed range to simulate different icing patterns. By combining probability density field and inverse transformation sampling to determine the attachment position, the icing path is sampled by arc length, and the curvature of each sampling point is calculated. An icing probability density field is constructed by combining the angle between wind direction and the power line normal, and wind speed. After normalization, the probability density function is obtained, and the cumulative distribution function is calculated. By generating uniform random numbers and performing inverse transformation sampling, the attachment position of the icicle is determined, making the icicles denser in high-curvature, windward regions. Finally, the windward deflection morphology is simulated, using the cross product of the wind direction vector and the gravity vector as the deflection axis. The deflection angle is determined according to the wind speed, and a rotation transformation is applied to the icicle model, presenting a windward tilted state. Therefore, this invention avoids the problems of randomized icicle attachment and idealized morphology, significantly improving the sample realism. Simultaneously, through parameterized generation and probability sampling, it balances sample diversity and generation efficiency.

[0036] S105 causes icicles to attach to the path of the icing defect generation on the power transmission line, thus obtaining a virtual scene of the icing defect.

[0037] This invention precisely combines parameterized icicles with icing paths through physical constraints and spatial positioning to construct a virtual scene with realistic icing characteristics. In practical applications, based on the determined attachment position and deflection shape of the icicles, fixed constraints can be used in Unreal Engine to bind the icicles to power line path nodes. One end of the constraint is connected to the center of the top of the icicle, and the other end is bound to a sampling point on the power line. Appropriate linear and rotational stiffness are set to ensure stable attachment of the icicles without abnormal displacement. Simultaneously, a global gravity field is activated to cause the icicles to exhibit slight natural drooping deformation, conforming to realistic mechanical effects. Finally, these constraints are integrated to form a virtual scene of icing defects. This invention completes the simulation construction of icing defects, obtaining a high-fidelity virtual scene that closely resembles actual working conditions, providing core materials for subsequent sample data conversion.

[0038] S106, convert the virtual scene of icing defects into usable sample data.

[0039] This invention transforms visual information from virtual scenes into a format recognizable and trainable by computer vision models, while ensuring the standardization and usability of sample data. In practical applications, Unreal Engine's built-in scene capture tools can be used. By adjusting the tool's pose to ensure the lens is aligned with the icing defect area, and setting the image resolution and storage format, images of icing scenes can be captured in batches. Simultaneously, the engine's blueprint function can automatically generate sample labels, annotating information such as the location, shape, and thickness of the icing. The image data and label data are stored accordingly to form a standardized sample dataset, suitable for subsequent detection model training. This invention completes the transformation from simulated scenes to practical samples. The generated samples possess advantages such as high realism, strong diversity, and accurate labeling, and can be directly used for model training, solving the traditional problem of sample acquisition.

[0040] This invention uses Unreal Engine as its core simulation platform, forming a closed-loop technical process from model construction, scene building, environmental control, icing simulation to sample conversion. It breaks through the limitations of traditional simulations that only focus on geometric morphology, achieving high-fidelity modeling of power equipment through a combination of 3D scanning and image acquisition. It calculates the actual sag path of transmission lines based on the parabolic approximation method as the icing benchmark, integrates probability density fields and inverse transformation sampling to ensure that the icing location conforms to natural laws, and combines wind speed parameters to simulate the windward deflection of icicles. Simultaneously, it enriches sample diversity through the control of multiple environmental parameters, ultimately generating standardized, high-fidelity icing sample data. Balancing physical realism and efficient generation, it deeply integrates mechanical calculations, probability statistics, and virtual simulation technology, specifically addressing the technical pain points of traditional icing sample acquisition difficulties, poor realism, and insufficient diversity.

[0041] like Figure 2 The diagram shown illustrates another flowchart of the physical simulation sample generation method for icing defects in power transmission lines according to the present invention, which may include: S201, Model Building.

[0042] Point cloud data and image data of a section of a 220kV transmission line were collected using a 3D scanner and a visible light camera. Then, based on the point cloud and images, the 3D models of the transmission tower, transmission line and other connecting equipment were accurately reconstructed in 3ds Max 3D modeling software. At the same time, icicles of various shapes were modeled.

[0043] It should be noted that 3D scanners can accurately capture the spatial geometric information of various components of transmission lines, generating high-density point cloud data. This point cloud data can completely reconstruct the three-dimensional outline, size proportions, and spatial relationships of transmission towers, transmission lines, and connecting equipment, solving the problem of insufficient geometric accuracy in traditional modeling. Visible light cameras simultaneously acquire scene image data, recording not only the surface texture and color of equipment but also providing a realistic reference for subsequent model texture mapping and material restoration, compensating for the lack of appearance details in point cloud data. Using 3ds Max as the core tool, the acquired point cloud data and image data serve as dual bases for the 3D model reconstruction of the core components of the transmission line. During the modeling process, the point cloud data is used as a benchmark to accurately calibrate the tower structure, crossarm angle, transmission line routing and suspension status, as well as the external dimensions and installation positions of connecting equipment such as insulators and hardware, ensuring geometric consistency between the model and the actual transmission line. Meanwhile, in response to the simulation requirements of icing defects, we conducted modeling work on various forms of ice columns separately. Combining the actual growth characteristics of icing on transmission lines, we designed ice column models with different thicknesses, lengths, curvatures, and attachment patterns, and built a basic ice column model library.

[0044] In terms of scene adaptability, the synchronous acquisition of visible light image data provides a foundation for the detailed reproduction of the constructed model. Subsequent optimization of the appearance and texture can be achieved through material mapping and parameter adjustments, ensuring that the model can quickly adapt to scene construction requirements after being imported into Unreal Engine. Simultaneously, the model's high-precision geometric features allow for perfect matching with subsequent operations such as collision volume settings and physical property binding, avoiding physical simulation anomalies caused by model structural issues and ensuring the smooth progress of the entire virtual simulation process.

[0045] S202, building a virtual scene of power transmission lines in Unreal Engine.

[0046] After modeling was completed, the 3D models of the power equipment and icicles were imported into Unreal Engine. The PBR (Physically Based Rendering) material system was used to add realistic-looking materials to all the power equipment and icicle models. In practical applications, the material textures for the power equipment can be obtained from visible light photographs, and the materials for the icicle models can be constructed by setting parameters such as refractive index and roughness. Complex collision volumes matching the model's own structure were then set for all power equipment components in the virtual scene.

[0047] It's important to note that importing the completed 3D models of the power equipment and icicles into Unreal Engine ensures the complete preservation of their geometric structure, providing a core framework for subsequent scene construction. During material optimization, a PBR material system is used to guarantee scene realism. By simulating real physical optics, the texture of objects is reproduced, and differentiated material construction methods are employed for different models. For example, the material textures of the power equipment are obtained from previously collected visible light photographs, maximizing the reproduction of surface texture, color, and wear marks, achieving a high degree of consistency with the real equipment in appearance. The icicle model's material is constructed by precisely setting core parameters such as refractive index and roughness. The refractive index determines the icicle's refraction of light, while roughness affects the degree of surface reflectivity. Finally, complex collision volumes matching the model's structure are set for all power equipment components. The design of these collision volumes closely matches the equipment's outline and structural characteristics, preventing deviations in subsequent physical simulations due to simplified collision volumes, and ensuring the scene possesses the basic conditions for physical interaction.

[0048] S203, Set virtual scene environment parameters.

[0049] In Unreal Engine, the built-in weather system can be used to change the weather to snow, covering the scene with snow. Its built-in lighting system can simulate lighting conditions at any time of day within 24 hours, enriching the diversity of the sample's scene environment.

[0050] As an example in this embodiment, virtual scene environment data can be set, including global wind direction. With global wind speed Maximum wind speed .

[0051] It should be noted that the engine's built-in weather system can switch the scene weather to snow and cover it with snow, accurately recreating the typical natural scene of icing phenomena. This ensures that the simulation environment is consistent with real icing conditions, providing a reasonable environmental background for the subsequent generation of icing defects. At the same time, the engine's built-in lighting system can flexibly simulate lighting conditions at any time within 24 hours. By adjusting parameters such as lighting angle, intensity, and color temperature, the limitations of a single lighting scene can be broken, enriching the environmental diversity of the samples and adapting to the diverse needs of subsequent sample shooting.

[0052] S204 sets the physical properties of the power transmission line.

[0053] Set the weight per unit length of the transmission line and horizontal tension It is used to calculate its natural sag curve as the path for the generation of icing defects.

[0054] It should be noted that the weight per unit length directly reflects the mass distribution characteristics of the transmission line itself, while the horizontal tension corresponds to the tensile force acting on both ends of the transmission line in a suspended state, and is used to calculate the natural sag curve of the transmission line. By establishing a strong correlation between physical parameters and the icing defect generation path, that is, by substituting the set weight per unit length and horizontal tension into subsequent engineering calculation models such as the parabolic approximation method, the natural sag shape of the transmission line under its own weight can be solved. This sag curve is then clearly defined as the generation path of the icing defect, ensuring that the icicle model can be accurately attached to the actual spatial position of the transmission line, and avoiding the disconnect between the icing path and the actual shape of the transmission line.

[0055] S205, simulates the generation of icing defects.

[0056] Step S205 defines an automatic icing defect generation algorithm based on physical simulation. This algorithm takes a power equipment model, various icicle models, and set environmental parameters as input, and automatically generates diverse icing defects along the transmission line path through the following sub-algorithm steps. For example... Figure 3 The diagram shown illustrates a process for generating icing defects on power transmission lines, which may include: (5.1) Physics-based path calculation.

[0057] Select the target power line in the virtual scene and obtain the three-dimensional coordinates of the two suspension endpoints. and As input, calculate the sag path of the target transmission line.

[0058] First, calculate the horizontal distance between the two suspension endpoints based on their three-dimensional coordinates. and vertical height difference :

[0059]

[0060] To simplify calculations, a parabolic approximation method commonly used in engineering is employed in a local two-dimensional plane, based on the set weight per unit length of the transmission line. Horizontal tension As a physical parameter, the maximum sag at the midpoint of the transmission line is estimated. :

[0061] in, It refers to the horizontal span between the two suspension ends of the power transmission line.

[0062] Starting point Establish a local coordinate system with the origin and sample at equal intervals along the horizontal span direction. The point. For the first point. Each sampling point, its local coordinates Corresponding height The equation of the parabola is given by the following:

[0063] This formula describes the shape of the parabola and simulates the sag caused by its own weight. It calculates all path points... Transforming from the local coordinate system back to the world coordinate system of the 3D scene yields an ordered set of 3D points. The calculated sequence of theoretical path points is used to automatically generate a smooth 3D spline curve using the spline curve component in Unreal Engine, which serves as the path for generating icing defects.

[0064] (5.2) Parameterized generation of icicle model.

[0065] In the blueprint control program, models are randomly selected from the list of icicle models for parametric generation. Within a fixed range, the 3D scaling, rotation, and transparency of the icicles are randomly set to simulate different forms of icing.

[0066] (5.3) Calculate the pose of the icicle model.

[0067] The path of icing defect generation is sampled, with arc length as the data. For each sampling point, calculate the parameters. curvature By combining virtual scene weather data (wind direction, wind speed) and path curvature, an icing probability density field is generated. :

[0068] in, It is the angle between the wind direction and the normal direction of the power transmission line. and All are weighting coefficients. Sampling points The curvature. For Normalization is performed to make it consistent with the total arc length. The integral over the given area equals 1, forming a standard probability density function. Calculate the icing probability density field. The cumulative distribution function (CDF) :

[0069] in, It is a function that monotonically increases from 0 to 1, establishing the relationship between arc length and probability. A unique mapping for the interval. The total number of icicles to be generated is determined based on the snow level in the virtual scene. (Increases with snow level). In Generate within the interval Given a uniformly distributed set of random numbers, for each random number... The corresponding position on the path is calculated by inverse transform sampling. .

[0070]

[0071] in, It is the inverse function of CDF. Therefore, depending on the set snow level, a greater number of icicle models will be generated under heavy snow conditions. Furthermore, based on the ice cover distribution probability density function, icicles can be made denser in high-probability areas and sparser in low-probability areas.

[0072] To simulate the impact of virtual wind on ice growth, virtual weather data was acquired to determine the deflection axis and deflection angle of the ice columns. Deflection axis Wind direction vector With gravity vector The direction of the cross product, i.e. Deflection angle Then it depends on the wind speed Decision, that is The calculated value around the deflection axis Rotation deflection angle The angle change was applied to the icicle model, making it appear to deflect in the wind.

[0073] (5.4) Add physical constraints.

[0074] A physics component was enabled for the icicle, setting its motion type to "Physically Affected." A mass of 0.1-0.5 kg / m was linearly assigned according to its 3D scaling value, and a matching convex polygon collider was added. Then, a fixed constraint was added at the contact point between the icicle and the power line, binding one end of the constraint to the center of the icicle's top and the other end to a node along the power line's path. A rigid connection was achieved by configuring a linear stiffness of 10000 N / m and a rotational stiffness of 8000 N·m / rad. A global gravity field was then enabled, limiting the icicle's horizontal displacement while allowing a natural vertical deformation of 0.1-0.3 mm / m. Finally, batch configuration was performed through a control program to ensure the icicle did not detach and that the physical effects conformed to real-world mechanical laws.

[0075] S206, shooting virtual samples.

[0076] Samples were captured using the Scene Capture 2D tool in Unreal Engine. The image resolution was set to 1920×1080 in the Blueprint Control program, and the storage format was PNG. The spatial position, rotation angle, pitch angle, and roll angle of the scene capture tool were adjusted to ensure the captured image was aligned with the icing defects. Weather conditions in the scene could be switched, and lighting parameters adjusted to enrich the virtual scene environment and increase the diversity of the generated samples.

[0077] In this embodiment, the Unreal Engine's built-in Scene Capture 2D tool is selected as the core acquisition method. It possesses precise scene capture capabilities and can adapt to the standardized acquisition requirements of virtual samples. By adjusting the spatial position, rotation angle, pitch angle, roll angle, and other pose parameters of the scene capture tool, the image is precisely aligned with the icing defect area, preventing the image from deviating from the core target and ensuring that each sample clearly presents the morphology, distribution, and other key features of the icing defects. Simultaneously, the Unreal Engine's scene control capabilities are fully utilized; by switching weather and adjusting lighting parameters, icing scenes under different environmental conditions are replicated, breaking the limitations of single-environment samples and achieving sample diversity expansion from the acquisition stage.

[0078] like Figure 4 The diagram shown is a schematic of a physical simulation sample generation system for icing defects in power transmission lines according to the present invention, which may include: The model module is used to build power equipment models and various types of icicle models, respectively. The virtual module is used to import power equipment models and icicle models into Unreal Engine to obtain a virtual scene of power transmission lines; The parameter module is used to set the environmental parameters of the virtual scene of the power transmission line; The calculation module is used to calculate the icing defect generation path of the transmission line based on the power equipment model, icicle model and environmental parameters, and to parameterize the icicle model. Combining the probability density field and inverse transformation sampling, it determines the attachment position of the parameterized icicle on the transmission line and makes the icicle present a windward deflection shape. Combined with the module, it is used to make icicles adhere to the path of icing defects in power transmission lines, thereby obtaining a virtual scene of icing defects. The conversion module is used to convert virtual scenes of icing defects into usable sample data.

[0079] It should be noted that, in the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of each block is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple blocks may be combined or integrated into another device, or some features may be ignored or not executed. The modules described as separate components may or may not be physically separated. The components shown as modules may be one or more physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] Furthermore, in the various embodiments of the present invention, the modules can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated unit described above can be implemented in hardware or as a software functional unit.

[0081] This invention also provides an electronic device, which may include one or more processors, memory, and communication interfaces.

[0082] The memory, communication interface, and processor are coupled together. For example, the memory, communication interface, and processor can be coupled together via a bus.

[0083] The communication interface is used for data transmission with other devices. The memory stores computer program code. This computer program code includes computer instructions, which, when executed by the processor, cause the electronic device to perform the steps of the aforementioned method for generating physical simulation samples of icing defects in power transmission lines.

[0084] The processor can be a processor or controller, such as a Central Processing Unit (CPU), a general-purpose processor, a Digital Signal Processor (DSP), an Application-Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with this disclosure. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The processor can be used to support an electronic device in performing the method steps provided in the above embodiments.

[0085] The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. These buses can be categorized as address buses, data buses, control buses, etc.

[0086] This invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the steps of the above-described method for generating physical simulation samples of icing defects in power transmission lines.

[0087] The computer-readable storage media involved in this invention include random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage media known in the art.

[0088] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

Claims

1. A method for generating physical simulation samples of icing defects in power transmission lines, characterized in that, include: Construct models of electrical equipment and various forms of icicles respectively; The power equipment model and icicle model were imported into Unreal Engine to obtain a virtual scene of the power transmission line; Set the environmental parameters for the virtual scene of the power transmission line; Based on the power equipment model, icicle model and environmental parameters, the generation path of icing defects on the transmission line is calculated, and the icicle model is parameterized. Combining the probability density field and inverse transformation sampling, the attachment position of the parameterized icicle on the transmission line is determined, and the icicle is made to present a windward deflection shape. By attaching icicles to the path of icing defects in the power transmission line, a virtual scene of icing defects is obtained. The virtual scenario of icing defects is transformed into usable sample data.

2. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, Methods for constructing power equipment models include: Point cloud data and image data of the power transmission line were collected using a 3D scanner and a visible light camera. Three-dimensional models of transmission towers, transmission lines, and connecting equipment are reconstructed based on point cloud data and image data.

3. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, The method for importing power equipment models and icicle models into Unreal Engine to obtain a virtual scene of a power transmission line includes: Import the power equipment model and the icicle model into Unreal Engine, add materials to the power equipment model and the icicle model, and then set the collision volume for the power equipment model and the icicle model.

4. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, The method for setting environmental parameters for a virtual transmission line scenario includes: Set the weather conditions in the environment to snow and cover it with snow; Simulate lighting conditions at any time of day using the lighting system in Unreal Engine; Set environmental data for wind speed and wind direction.

5. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, The method for calculating the generation path of icing defects in transmission lines includes: Estimate the maximum sag at the midpoint of the transmission line : in, This refers to the weight per unit length of the power transmission line. The horizontal tension of the transmission line. This refers to the horizontal span between the two suspension ends of the power transmission line. Starting point of the target transmission line suspension endpoint Establish a local coordinate system with the origin and sample at equal intervals along the horizontal span direction. For each sampling point, calculate the height corresponding to the local coordinates. : in, For the first The x-coordinate of the local coordinates of each sampling point; The vertical height difference between the two suspension ends of the target transmission line; All calculated path points are transformed from the local coordinate system to the world coordinate system of the 3D scene to obtain a 3D point set; the x-coordinate of the path point is the x-coordinate of the local coordinate of the sampling point, and the y-coordinate is the corresponding height. Based on the 3D point set, a 3D spline curve is generated using the spline curve component in Unreal Engine, which serves as the path for generating icing defects.

6. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, The method for determining the attachment location of parameterized icicles on power transmission lines by combining probability density fields and inverse transform sampling includes: The path of icing defect generation is sampled, with arc length as the data. Calculate the sampling points for each parameter. curvature ; Calculate the ice accretion probability density field : in, It is the angle between the wind direction and the normal direction of the power transmission line. and All are weighting coefficients. For global wind speed; For the ice accretion probability density field Normalization is performed to obtain the standard probability density function; By combining the standard probability density function, the cumulative distribution function of the icing probability density field is calculated; Based on the snow level in the virtual scene, determine the total number of icicles to be generated and generate uniformly distributed random numbers accordingly. For each random number, sample it by inverse transformation of the cumulative distribution function of the ice accretion probability density field and calculate the corresponding position of the random number on the ice accretion defect generation path.

7. The method for generating physical simulation samples of icing defects in transmission lines according to claim 1, characterized in that, The method for causing icicles to deflect in the wind includes: Determine the deflection axis and deflection angle of the icicle respectively; By changing the angle of rotation around the deflection axis, the icicle is made to deflect in the wind.

8. A physical simulation sample generation system for icing defects in power transmission lines, characterized in that, include: The model module is used to build power equipment models and various types of icicle models, respectively. The virtual module is used to import power equipment models and icicle models into Unreal Engine to obtain a virtual scene of power transmission lines; The parameter module is used to set the environmental parameters of the virtual scene of the power transmission line; The calculation module is used to calculate the icing defect generation path of the transmission line based on the power equipment model, icicle model and environmental parameters, and to parameterize the icicle model. Combining the probability density field and inverse transformation sampling, it determines the attachment position of the parameterized icicle on the transmission line and makes the icicle present a windward deflection shape. Combined with the module, it is used to make icicles adhere to the path of icing defects in power transmission lines, thereby obtaining a virtual scene of icing defects. The conversion module is used to convert virtual scenes of icing defects into usable sample data.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for generating physical simulation samples of icing defects in transmission lines as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for generating physical simulation samples of icing defects in transmission lines as described in any one of claims 1-7.

Images