A method, system, equipment, and medium for managing tree obstructions along power transmission lines.

Abstract

This invention relates to the field of transmission line management technology, and discloses a method, system, equipment, and medium for managing tree obstructions along transmission lines. The method involves acquiring sparse sample data for each section of the transmission line; inputting conditional variables constructed from the sparse sample data and real sample data constructed from the tree data of the sparse sample data into a trained improved generative adversarial network (GAN) to generate sample data; the generator loss function of the improved GAN includes a first tree height constraint function determined by soil moisture, a second tree height constraint function determined by tree age, a third tree height constraint function determined by ground slope, a tree tilt angle constraint function, and / or a tree growth rate constraint function; and performing Monte Carlo simulation on the tree sample dataset composed of real and generated sample data to obtain a tree-line contact time set, thereby generating a corresponding tree obstruction management strategy. This method improves the efficiency and effectiveness of tree obstruction management along transmission lines.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission line management, and in particular to a power transmission line tree barrier management method, system, device and medium. BACKGROUND

[0002] At present, the power transmission line of the power distribution network usually passes through forest, green belt and other vegetation dense areas. In these areas, the continuous growth of trees may cause them to gradually approach or contact the power transmission line, bringing safety hazards such as short circuit, tripping, equipment damage, and even large-scale power outage accidents in severe cases.

[0003] At present, the power transmission line tree barrier management mainly adopts artificial fixed periodic inspection or equipment monitoring. However, the above methods have significant limitations: the cycle of artificial fixed periodic inspection lacks scientific basis and cannot dynamically adapt to the growth law of different tree species and meteorological changes, which easily leads to over-pruning, delayed pruning, resource waste and other problems. Second, relying on monitoring equipment for monitoring, there are difficulties in equipment installation and power supply, especially in areas where vegetation is abundant on both sides of the power transmission line, unevenly distributed, and the terrain spans lakes and mountains and other diverse areas, a large number of sensors and monitoring equipment need to be deployed to collect a large amount of field data, resulting in high equipment installation and maintenance costs.

[0004] Therefore, how to improve the efficiency, effectiveness and scientificity of power transmission line tree barrier management has become a technical problem to be solved by those skilled in the art. SUMMARY

[0005] The present application provides a power transmission line tree barrier management method, system, device and medium to solve the technical problem of how to improve the efficiency, effectiveness and scientificity of power transmission line tree barrier management, and achieve the effect of improving the efficiency, effectiveness and scientificity of power transmission line tree barrier management.

[0006] In a first aspect, the present application provides a power transmission line tree barrier management method, which comprises:

[0007] Dividing a target power transmission line into sections and obtaining sparse sample data of each section of the power transmission line, wherein the sparse sample data includes tree data, terrain data and environmental data;

[0008] Constructing a conditional variable according to the sparse sample data, constructing a real sample data according to the tree data, inputting the conditional variable and the real sample data into an improved generative adversarial network trained, and obtaining generated sample data; the generator loss function of the improved generative adversarial network at least includes a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by tree age, a tree inclination angle constraint function and / or a tree growth rate constraint function;

[0009] Monte Carlo simulation was performed on the tree sample dataset composed of the real sample data and the generated sample data to obtain the tree-line contact time set corresponding to each section of the transmission line.

[0010] Based on the tree-line contact time set, a tree obstacle management strategy is generated for each section of the transmission line.

[0011] Preferably, the step of dividing the target transmission line into sections and obtaining sparse sample data for each section of the transmission line includes:

[0012] The trees around the target transmission line are divided according to their growth patterns to obtain several tree species groups;

[0013] Based on the tree species group and the year of planting of the trees surrounding the target transmission line, the target transmission line is divided into several sections of transmission line.

[0014] Tree data for each section of the transmission line is collected, and the tree data includes at least: initial tree height, tree tilt angle, tree growth rate, initial tree age, and horizontal distance from the tree line;

[0015] Collect topographic data for each section of the transmission line, including at least the ground slope.

[0016] Environmental data for each section of the transmission line is collected, and the environmental data includes at least: soil moisture, light intensity and annual precipitation.

[0017] Based on the tree data, the terrain data, and the environmental data, sparse sample data is constructed.

[0018] Preferably, the step of constructing conditional variables based on the sparse sample data and constructing real sample data based on the tree data includes:

[0019] Conditional variables are constructed based on the ground slope, soil moisture, light intensity, annual precipitation, and initial tree age;

[0020] Real sample data is constructed based on the initial height of the trees, the tilt angle of the trees, the growth rate of the trees, and the horizontal distance of the tree line.

[0021] Preferably, the first tree height constraint function is set to reflect the relationship between the tree height and the ground slope;

[0022] The second tree height constraint function is set to reflect the relationship between the tree height and the soil moisture;

[0023] The tree tilt angle constraint function is set such that the tree tilt angle should be less than the sum of the ground slope and the safety margin;

[0024] The tree growth rate constraint function is set such that the tree growth rate should be determined by both the light intensity and the annual precipitation.

[0025] Preferably, the discriminator loss function of the improved generative adversarial network includes at least a gradient penalty term, a deviation of the true sample score from the expected value term, and a deviation of the generated sample score from the expected value term;

[0026] The gradient penalty term is set to reflect the discriminator's expectation of the gradient of the interpolated sample data, and the interpolated sample data is set to be obtained by interpolation between the real sample data and the generated sample data;

[0027] The actual sample score deviation from expectation term is set to reflect the degree of deviation between the discriminator’s first score for the actual sample data and the expected score for the actual sample data;

[0028] The generated sample score deviation from expectation term is set to reflect the degree of deviation between the discriminator's second score for the generated sample data and the expected score for the generated sample data.

[0029] Preferably, the step of performing Monte Carlo simulation on the tree sample dataset composed of the real sample data and the generated sample data to obtain the tree-line contact time set corresponding to each transmission line segment includes:

[0030] Based on the real sample data and the generated sample data, a tree sample dataset is obtained. Based on the initial height of the trees and the tree growth rate in the tree sample dataset, a tree growth function is constructed.

[0031] Based on the tree growth function, a tree line relationship model is constructed, and based on the tree line relationship model, the current tree line distance is obtained;

[0032] A disturbance factor is introduced into the current tree line distance to obtain the current tree line disturbance distance, and a tree line contact determination model is constructed based on the current tree line disturbance distance and a pre-set safety radius;

[0033] Based on the tree line contact determination model, tree line contact determination is performed on each tree sample data in the tree sample dataset to obtain the tree line contact determination result for each tree sample data in the tree sample dataset.

[0034] Based on the tree-line contact determination results, the tree-line contact time set corresponding to each section of the transmission line is obtained.

[0035] Preferably, generating a tree obstacle management strategy for each transmission line segment based on the tree-line contact time set includes:

[0036] Cluster analysis was performed on the tree-line contact time set to obtain several cluster centers;

[0037] Based on each cluster center and a pre-set contact time safety margin, a tree barrier management strategy is generated for each transmission line segment, wherein the tree barrier management strategy includes at least a tree pruning time series.

[0038] Secondly, the present invention also provides a transmission line tree obstacle management system to implement the above-described transmission line tree obstacle management method. The system includes: a data acquisition module, an improved generative adversarial network module, a Monte Carlo simulation module, and a tree obstacle management strategy generation module.

[0039] The data acquisition module is used to divide the target transmission line into sections and acquire sparse sample data of each section of the transmission line. The sparse sample data includes tree data, terrain data and environmental data.

[0040] The improved generative adversarial network module is used to construct conditional variables based on the sparse sample data, construct real sample data based on the tree data, and input the conditional variables and the real sample data into the trained improved generative adversarial network to obtain generated sample data; the generator loss function of the improved generative adversarial network includes at least a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by the tree age, a tree tilt angle constraint function, and / or a tree growth rate constraint function;

[0041] The Monte Carlo simulation module is used to perform Monte Carlo simulation on the tree sample dataset composed of the real sample data and the generated sample data to obtain the tree-line contact time set corresponding to each section of the transmission line.

[0042] The tree obstacle management strategy generation module is used to generate a tree obstacle management strategy corresponding to each section of the transmission line based on the tree-line contact time set.

[0043] Thirdly, the present invention also provides a computer device, the computer device including a memory, a processor and a transceiver, which are connected to each other via a bus; the memory is used to store a set of computer program instructions and data, and to transmit the stored data to the processor, the processor executing the computer program instructions stored in the memory to perform the above-described method for managing tree obstructions on power transmission lines.

[0044] Fourthly, the present invention also provides a computer-readable storage medium storing a computer program that, when executed, implements the above-described method for managing tree obstructions along power transmission lines.

[0045] This application provides a method, system, device, and medium for managing tree obstructions along power transmission lines. Compared to the prior art, the beneficial effects of the embodiments of this application are as follows:

[0046] The tree obstacle management method for transmission lines disclosed in this application generates a large amount of generated sample data based on sparse sample data. It does not rely on large-scale sensor deployment or require full historical growth data. It is particularly suitable for transmission lines traversing forest areas, lake areas, and mountainous areas where monitoring is difficult, effectively reducing the management cost and maintenance burden of tree obstacles. By modeling the tree growth process and simulating the behavior of trees in contact with the line, potential contact risks can be predicted in advance. It no longer relies on fixed-period pruning patterns and supports "on-demand pruning." By replacing blind pruning with scientific prediction, it achieves refined management and a shift from passive maintenance to proactive management, effectively saving maintenance resources and reducing unnecessary labor costs and pruning frequency. At the same time, it provides early warning of high-risk tree obstacles, significantly reducing the occurrence of line tripping and power outages, ensuring the stable operation of the power system, and improving power supply reliability. It has good social benefits and promotional value. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the steps of a method for managing tree obstructions on power transmission lines according to a preferred embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram illustrating the result of dimensionality reduction visualization of real sample data and generated sample data using principal component analysis, provided in a preferred embodiment of the present invention.

[0049] Figure 3 This is a schematic diagram of the positional relationship of tree lines provided in a preferred embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram of the change curve of the contact probability of a tree with the number of simulations during multiple Monte Carlo simulations, provided by a preferred embodiment of the present invention.

[0051] Figure 5 This is a schematic diagram of the structure of a power transmission line tree obstacle management system provided in a preferred embodiment of the present invention;

[0052] Figure 6 This is an internal structural diagram of the computer device in an embodiment of the present invention;

[0053] Figure label:

[0054] 1-Data acquisition module, 2-Improved generative adversarial network module, 3-Monte Carlo simulation module, 4-Tree obstacle management strategy generation module. Detailed Implementation

[0055] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. The embodiments are provided for illustrative purposes only and should not be construed as limiting the invention. The accompanying drawings are for reference and illustration only and do not constitute a limitation on the scope of protection of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of this invention. In the description of this invention, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0056] In the description of this invention, it should be noted that, unless otherwise expressly specified and limited, the term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0057] In the description of this invention, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0058] Please see Figure 1 The diagram illustrates the steps of a method for managing tree obstructions along power transmission lines. In an embodiment of the present invention, a method for managing tree obstructions along power transmission lines is provided, the method comprising:

[0059] S1. Divide the target transmission line into sections and obtain sparse sample data for each section. The sparse sample data includes tree data, terrain data, and environmental data. In a preferred embodiment of this application, a method for managing tree obstacles on transmission lines suitable for data-sparse scenarios is disclosed. The section division of the target transmission line in this application is mainly based on the planting year and tree species of the trees. Factors that significantly affect the height of the trees include tree age, tree species, and climatic conditions. The age and species of trees around the transmission line generally have regional similarities. Therefore, based on the tree species and planting year of the target transmission line, the target transmission line is divided into several sections. Regarding the tree species as a basis, different species of trees are generally divided into different sections. However, some tree species, although different, have similar growth patterns and can be considered as the same species. Therefore, in this application, trees are divided into different tree species groups according to their growth patterns. Based on the planting year of trees, areas where the difference in planting year is less than or equal to a preset year threshold are divided into the same segment. The specific preset year threshold needs to be determined according to the type of tree and the climate conditions of the area. For example, if the trees are fast-growing and the climate conditions of the area are very suitable for the growth of the trees, and according to historical data, the trees can grow an average height of 3m in one year, then the trees in this area belong to the category of fast-growing trees. In this case, the preset year threshold needs to be selected as a small value, such as 1, so as to divide trees of different planting years into different segments.

[0060] Based on the planting year and tree species of the trees surrounding the target transmission line, this application divides the target transmission line into several sections. Within the same area of ​​the transmission line, the trees have similar original height data and growth curves, which improves the accuracy of subsequent sample generation data based on sparse sample data.

[0061] Furthermore, random sampling is conducted on trees surrounding the transmission line to collect tree data for each section of the transmission line. The tree data includes at least the initial height of the trees, the angle of inclination of the trees, the growth rate of the trees, the initial age of the trees, and the horizontal distance from the tree line. Topographic data of the transmission line corridor is collected through geographic information systems, digital elevation models, remote sensing data, or historical data. The topographic data includes at least the ground slope. Environmental data for each section of the transmission line is collected, including at least soil moisture, light intensity, and annual precipitation. Finally, sparse sample data for each section of the transmission line is constructed from the tree data, topographic data, and environmental data. This method can obtain initial sparse sample data without deploying a large number of sensors, which is particularly suitable for transmission line areas that traverse forest areas, lake areas, mountains, and other areas where monitoring is difficult, effectively reducing management costs and maintenance burden.

[0062] S2. Construct conditional variables based on the sparse sample data, construct real sample data based on the tree data, and input the conditional variables and the real sample data into the trained improved generative adversarial network to obtain generated sample data. In a preferred embodiment of this application, the generator loss function of the improved generative adversarial network includes at least a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by the tree age, a tree tilt angle constraint function, and / or a tree growth rate constraint function. The discriminator loss function of the improved generative adversarial network includes at least a gradient penalty term, a real sample score expectation term, and a generated sample score expectation term. In a preferred embodiment of this application, a small amount of sparse sample data is collected. To ensure the accuracy of the subsequent tree line contact analysis results, a large amount of sample data is generated using the improved generative adversarial network. The generation process of the improved generative network in this application not only needs to maintain the similarity between the generated sample data and the sparse sample data, but also needs to conform to basic physical and ecological laws. Therefore, this application introduces conditional variables and constructs a physical rule-guided loss function so that the generator learns the data distribution that conforms to real laws, effectively expands the small amount of structured data, and improves the reliability of subsequent simulation, prediction, and evaluation tasks. Specifically, this application selects ground slope, soil moisture, light intensity, annual precipitation and initial tree age to construct conditional variables, and uses the initial tree height, tree tilt angle, tree growth rate and horizontal distance of tree line as real sample data, i.e. reconstruction target.

[0063] To improve adversarial networks, condition variables are incorporated into the generator to control the direction of sample data generation, enhancing the physical interpretability of the generated sample data. Among these condition variables, ground slope ranges from [0, 60], soil moisture from [0.0, 1.0], light intensity from [0, 2000], annual precipitation from [800, 2000], and tree age from [1, 20]. After normalizing ground slope, soil moisture, light intensity, annual precipitation, and tree age, the condition vector composed of these condition variables can be represented as:

[0064]

[0065] Wherein, SLO represents the normalized ground slope, SMO represents the normalized soil moisture, RAD represents the normalized light intensity, PRC represents the normalized annual precipitation, and AGE represents the normalized tree age.

[0066] The original Generative Adversarial Network (GAN) uses Wasserstein distance as its loss function. Wasserstein distance is a metric that measures the distance between two probability distributions, making GAN training more stable and convergent. The expression for Wasserstein distance is:

[0067]

[0068] in, This represents the generator loss function of the original generative adversarial network. This represents the expected value of a random variable. A sample vector representing the real sample data. Represents a given condition vector and random variables The generator outputs the generated sample data. This indicates that the discriminator scores the generated sample data output by the generator.

[0069] However, the Wasserstein distance can only measure the degree of similarity between the generated sample data and the real sample data in terms of statistical distribution. In the preferred embodiment of this application, a physical rule loss function that conforms to reality is introduced into the generator so that the generated sample data conforms to the corresponding physical rules. The physical rule loss function includes at least: a tree tilt angle constraint function, a first tree height constraint function determined by terrain data, a second tree height constraint function determined by environmental data, a second tree height constraint function determined by tree age, and / or a tree growth rate constraint function.

[0070] Tree height should be adapted to the ground slope. When the slope is greater than 40°, tall trees generally will not grow, and the slope height is unstable, only supporting the growth of shrubs or small trees. Therefore, the first tree height constraint function is set to reflect the relationship between ground slope and tree height. The ground slope should be adapted to the tree height. The first tree height constraint function is expressed as follows:

[0071]

[0072] in, This represents the first tree height constraint function. This represents the original height of the trees used to generate the sample data.

[0073] Tree height should be consistent with soil moisture, and the tree height should be greater than or equal to 0. Higher trees generally grow in areas with high moisture content. Therefore, the first tree height constraint function is set to reflect the relationship between tree height and soil moisture. The tree height should be consistent with the second theoretical tree height that the soil moisture can support. The second tree height constraint function is expressed as follows:

[0074]

[0075] in, This represents the function that constrains the height of the second tree.

[0076] Tree height should be appropriate to tree age, and should not exceed the maximum height determined by tree age. Therefore, the second tree height constraint function is set to reflect the relationship between tree height and tree age. The tree height should be consistent with the second theoretical tree height determined by tree age. The second tree height constraint function is expressed as follows:

[0077]

[0078] in, This represents the function that constrains the height of the third tree.

[0079] The tree tilt angle constraint function is set so that the tree tilt angle should be less than the sum of the ground slope and the safety margin. The tree tilt angle constraint function is expressed as follows:

[0080]

[0081] in, This represents the tree tilt angle constraint function. This indicates the tree tilt angle used to generate the sample data. The safety margin is indicated by 10° in this application.

[0082] Tree growth rate is limited by both light intensity and precipitation. Therefore, the tree growth rate constraint function is set to be determined by both light intensity and annual precipitation, as follows:

[0083]

[0084] in, This represents the tree tilt angle constraint function. This indicates the rate of tree growth.

[0085] Therefore, the generator loss function of the improved adversarial generative network in this application is expressed as:

[0086]

[0087] in, This represents the generator loss function for improving generative adversarial networks. This represents the weight of the generator's physical rule loss function term. This represents the expected value of the generated sample data.

[0088] The goal of the discriminator in an adversarial generative network (GGN) is to award high scores to real sample data and low scores to generated sample data. The original GGN discriminator loss function is:

[0089]

[0090] in, This represents the discriminator loss function of the original generative adversarial network. Indicates the generation of sample data The distribution Represents real sample data The distribution This indicates that the discriminator generates sample data. Score them. This indicates that for all generated sample data The expected value for scoring. This indicates that the discriminator is effective against real sample data. Score them. This indicates that for all real sample data The expected value for scoring.

[0091] In a preferred embodiment of this application, to ensure the Lipshitz continuity of the discriminator and prevent training instability, a gradient penalty term, a deviation from expected real sample scores term, and a deviation from expected generated sample scores term are added to the discriminator loss function. The gradient penalty term aims to enable the discriminator to determine whether the structures of the real and generated sample data are reasonable. The gradient penalty term is expressed as follows:

[0092]

[0093] in, Represents the gradient penalty term. This represents the interpolated sample data obtained by interpolating between the real sample data and the generated sample data. This represents the gradient of the discriminator's score on the interpolated sample data. This represents the L2 norm, i.e., the Euclidean length. This represents the expectation of the gradient of the discriminator on the difference sample data.

[0094] The interpolated sample data is represented as follows:

[0095]

[0096] in, This represents a random number that is uniformly sampled between 0 and 1.

[0097] The loss function is equivalent to an optimization problem. The discriminator loss function aims to maximize the distinction between real sample data and generated sample data. For real sample data, it is desirable for the discriminator's score to be as close to 1 as possible. Therefore, the real sample score deviation from expectation term in this application is set to reflect the degree of deviation between the discriminator's first score for real sample data and the expected score for real sample data. When the discriminator's first score for real sample data deviates from 1, a corresponding loss will occur. For generated sample data, it is desirable for the discriminator's score to be as close to 0 as possible. Therefore, the generated sample score deviation from expectation term in this application is set to reflect the degree of deviation between the discriminator's second score for generated sample data and the expected score for generated sample data. When the discriminator's second score for generated sample data deviates from 0, a corresponding loss will occur. Therefore, the real sample score deviation from expectation term and the generated sample score deviation from expectation term constitute the structural scoring loss function, which is expressed as follows:

[0098]

[0099] in, Represents the structural scoring loss function. This indicates that the actual sample score deviates from the expected value. This indicates that the generated sample score deviates from the expected value.

[0100] Finally, the discriminator loss function is expressed as:

[0101]

[0102] in, Describes the discriminator loss function of the improved generative adversarial network. This represents the weight of the penalty gradient term. This represents the weights of the structural scoring loss function.

[0103] The weights, biases, and offset terms of the generator and discriminator of the improved adversarial network (AAN) are initialized. The ReLU (Rectified Linear Unit) activation function is chosen. For the discriminator's hidden layers, the last layer uses the Sigmoid activation function to output the discrimination result. Based on a training dataset consisting of historical condition variables and random noise, the generator and discriminator of the AAN are trained alternately until the convergence condition is met: the discriminator can no longer effectively distinguish between real and fake data, and the discriminator's output for both real and generated sample data is close to 0.5. At this point, the data distribution generated by the generator closely approximates the distribution of real sample data, and training terminates.

[0104] like Figure 2The diagram shows the results of dimensionality reduction visualization of real and generated sample data using principal component analysis. The large red dots represent the original 100 real sample data points, while the smaller blue dots represent the 5000 generated sample data points expanded from these 100 data points by the improved generative adversarial network. As shown in Figure 2, the overall distribution of the data generated by the improved generative network is quite consistent with the real sample data, and a relatively continuous simulated sample group forms near the distribution area of ​​the real sample data. This indicates that the improved generative adversarial network successfully learned the latent structural features of the real sample data while maintaining its characteristics, and possesses a certain degree of generalization ability.

[0105] S3. Perform Monte Carlo simulation on the tree sample dataset composed of the real sample data and the generated sample data to obtain the tree-line contact time set corresponding to each transmission line segment; in a preferred embodiment of this application, the conditional vector and real sample data are input into the trained improved generative adversarial network to obtain generated sample data, and the real sample data and generated sample data are mixed to obtain the tree sample dataset, which represents the initial parameter data of trees around the transmission line segment. In this application, the tree sample dataset is represented as:

[0106]

[0107] in, Represents a tree sample dataset, Indicates the first Tree sample data, This represents the index of tree sample data. This represents the total number of tree sample data.

[0108] like Figure 3 The diagram shows the positional relationship of the tree lines. Further, based on the initial tree height and tree growth rate in the tree sample dataset, a tree growth function is constructed, expressed as:

[0109]

[0110] in, express Time of the first The current tree height, Indicates the first The initial height of the trees. Indicates the first The growth rate of the trees Indicates a time index. Indicates the angle of tree tilt.

[0111] A coordinate system is established with the tree roots as the origin, and a tree-line relationship model is constructed. Based on the horizontal distance of the tree lines in each tree sample data set, the two-dimensional coordinates of the transmission line are represented as follows: ,in, Indicates the horizontal distance of the tree line. Indicates the height of the power transmission line.

[0112] like Figure 3 As shown, based on the current tree height and tilt angle obtained from the tree growth function, a tree model is drawn using modeling software. Then, assuming no tilt angle, the coordinates of the treetop are:

[0113]

[0114] Assuming the tree crown is elliptical, with the direction of the trunk as the major axis of the ellipse, and assuming the tree has no swaying angle, the coordinates of the tip of any branch are:

[0115]

[0116] in, Indicates the height of the tree trunk. Represents the semi-major axis of the ellipse. This represents the ratio between the major and minor semi-axis of an ellipse. This indicates the eccentric angle of the branch.

[0117] The trunk height of a tree typically refers to the height from the ground to the first major branch point of the trunk, or, when there are no obvious branches, to the base of the crown. It is one of the core characteristics of its growth. Its value is significantly related to the age and species of the tree, and is also influenced by environmental factors, but the core pattern is determined by both species and age. For tall trees, such as pines, firs, and spruces, the trunk is tall and the growth cycle is long. These are mostly tall evergreen trees with straight trunks and high branching points. Broadleaf tall trees, such as poplars, eucalyptus, sycamores, and camphor trees, have thick trunks, and the branching point gradually increases with age. Medium and low-growing trees have moderate trunk height and branch early. The trunk height of these trees is usually 5-15 meters, with lower branching points, and they focus more on the lateral expansion of the crown. Shrubs and small trees have short trunks, or even no obvious trunk. These trees do not tend to grow with tall trunks, and the trunk height after maturity is mostly below 5 meters, or even grow in clumps. During the juvenile stage of a tree, energy is preferentially allocated to the longitudinal growth of the trunk, resulting in fewer and shorter branches, and the trunk height increases most rapidly. In the youth stage, the tree begins to differentiate its energy allocation to trunk thickening and crown branching, but trunk height continues to increase. In the mature stage, the focus of growth shifts to trunk thickening, crown expansion, and reproduction, and trunk height growth slows considerably or even ceases. Therefore, this application determines the trunk height of trees based on age and species, fully considering the impact of different species data and growth stages on the position of any branch tip, thus improving the accuracy of the tree-line relationship model.

[0118] The ratio between the long and short axes can be determined by the tree species. For example, banyan, camphor, and mature oak trees have similar crown widths in all directions, so the ratio of the long axis to the short axis is close to 1:1. For mimosa and flame trees, which have wider lateral extensions and lower longitudinal heights, the long axis is horizontal, and the ratio of the long axis to the short axis is approximately 1.2:1 to 1.5:1. For pine, fir, and cypress trees, which are tall longitudinally and narrow laterally, the long axis coincides with the trunk, and the ratio of the long axis to the short axis is approximately 1.5:1 to 2:1.

[0119] Furthermore, the first distance between the treetop and the power transmission line, i.e., the first current tree-to-line distance, is calculated using the following formula:

[0120]

[0121] in, This indicates the distance to the first current tree line.

[0122] The second distance between the branch tip and the power transmission line, i.e., the second current tree-line distance, is calculated using the following formula:

[0123]

[0124] in, This indicates the distance to the second current tree line.

[0125] To address the swaying of trees caused by wind and unforeseen circumstances, a disturbance factor is introduced into the first and second current treeline distances, resulting in the first and second current treeline disturbance distances. These distances are expressed as follows:

[0126]

[0127]

[0128] in, This represents the disturbance factor. Indicates the current treeline disturbance distance. This indicates the second current treeline disturbance distance.

[0129] The disturbance factor is greater than 1. The specific value is determined based on the annual wind speed of the transmission line section. The higher the annual wind speed, the higher the disturbance factor.

[0130] Furthermore, based on the first current tree-line disturbance distance, the second current tree-line disturbance distance, and a pre-set safety radius, a tree-line contact determination model is constructed to determine whether a tree will come into contact with the transmission line. If either the first current tree-line disturbance distance or the second current tree-line disturbance distance is less than the pre-set safety radius, it indicates that there is a risk of contact between the tree and the transmission line.

[0131] The tree-line contact determination formula for the tree-line contact determination model is as follows:

[0132]

[0133] in, Indicates the safety radius. The number indicates the risk of exposure, with 1 indicating a risk of exposure and 0 indicating no risk of exposure.

[0134] Based on the tree-line contact determination model, tree-line contact determination is performed on each tree sample data in the tree sample dataset to obtain the tree-line contact determination result for each tree sample data in the tree sample dataset. If contact is found, the contact time is recorded to obtain the tree-line contact time set corresponding to each section of the transmission line. Based on the tree-line contact determination results, the tree-line contact probability of each area of ​​the transmission line is determined. The formula for calculating the tree-line contact probability is as follows:

[0135]

[0136] in, Indicates the first The probability of tree-line contact in a section of the transmission line. Indicates the first The number of simulations of tree-line contact events occurring in each section of the transmission line. Indicates the first The total number of simulations for each section of the transmission line. This indicates the section of the transmission line index.

[0137] S4. Based on the tree-line contact time set, generate a tree obstacle management strategy corresponding to each section of the transmission line; In this application, by performing cluster analysis on the tree-line contact time set, several cluster centers are obtained. The product of the contact time represented by each cluster center and the pre-set contact time safety margin is taken as a tree pruning time, generating a tree pruning time sequence, and obtaining the tree obstacle management strategy corresponding to each section of the transmission line.

[0138] In another preferred embodiment of this application, the expected tree-line contact time for each area's transmission line is calculated based on the tree-line contact probability and tree-line contact time set. The expected tree-line contact time is as follows:

[0139]

[0140] in, An index representing the trees where tree-line contact events have occurred. Indicates the first The duration of tree-line contact for the trees that experienced the tree-line contact event. This indicates the expected time for tree line contact.

[0141] The tree pruning cycle for each area of ​​the transmission line is obtained by multiplying the expected tree contact time and the pre-set safety margin for contact time. The trees around the transmission line in that area are pruned according to the tree pruning cycle.

[0142] In a preferred embodiment of this application, a 10 kV transmission line in a certain area is selected as the simulation object. This transmission line traverses a mixed forest and slope area, with a total length of approximately 8.2 kilometers. The tree species in the area are mainly moso bamboo, fir, and chinaberry. As per step S1, the line is divided into 5 regions. Since moso bamboo is the key focus in this region, this case study only uses moso bamboo as the research object. Sparse sample data of 100 trees on both sides of the transmission line are collected and extracted. An improved generative adversarial network is used to augment the sparse sample data of these 100 trees, resulting in a final sample of 5000 trees.

[0143] like Figure 4 The diagram shows the change in the contact probability of trees as the number of simulations increases during multiple Monte Carlo simulations. The horizontal axis represents the number of simulations, and the vertical axis represents the cumulatively calculated contact probability. Figure 4As the number of simulations gradually increases, the contact probability gradually stabilizes, indicating that under the current parameter settings, the simulation process has good convergence and the calculated contact probability has high reliability. The figure reflects the impact of the number of simulation samples on the accuracy of the results: when the number of simulations is small (e.g., n < 200), the curve fluctuates significantly, exhibiting a certain degree of randomness; however, when the number of simulations reaches a certain threshold (e.g., n > 1000), the change in contact probability slows down, indicating that the sample size is sufficient to support stable estimation.

[0144] In a preferred embodiment of the present invention, the target transmission line is divided into sections, and sparse sample data of each section is obtained. The sparse sample data includes tree data, terrain data, and environmental data. Conditional variables are constructed based on the sparse sample data, and real sample data is constructed based on the tree data. The conditional variables and real sample data are input into a trained improved generative adversarial network to obtain generated sample data. The generator loss function of the improved generative adversarial network includes at least a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by tree age, a tree tilt angle constraint function, and / or a tree growth rate constraint function. Monte Carlo simulation is performed on the tree sample dataset composed of real sample data and generated sample data to obtain the tree-line contact time set corresponding to each section of the transmission line. Based on the tree-line contact time set, a tree obstacle management strategy corresponding to each section of the transmission line is generated. The tree obstacle management method for transmission lines disclosed in this application generates a large amount of generated sample data based on sparse sample data. It does not rely on large-scale sensor deployment or require full historical growth data. It is particularly suitable for transmission lines traversing forest areas, lake areas, and mountainous areas where monitoring is difficult, effectively reducing the management cost and maintenance burden of tree obstacles. By modeling the tree growth process and simulating the behavior of trees in contact with the line, potential contact risks can be predicted in advance. It no longer relies on fixed-period pruning patterns and supports "on-demand pruning." By replacing blind pruning with scientific prediction, it achieves refined management and a shift from passive maintenance to proactive management, reducing the frequency of unnecessary pruning and effectively improving the efficiency of tree obstacle management for transmission lines. At the same time, it provides early warning of high-risk tree obstacles, significantly reducing the occurrence of line tripping and power outages, ensuring the stable operation of the power system, improving power supply reliability, and has good social benefits and promotional value.

[0145] Accordingly, such as Figure 5 The diagram shown illustrates the structure of a power transmission line tree barrier management system. Based on a power transmission line tree barrier management method, this embodiment of the invention also provides a power transmission line tree barrier management system that implements the power transmission line tree barrier management method disclosed in this embodiment. The system includes: a data acquisition module 1, an improved generative adversarial network module 2, a Monte Carlo simulation module 3, and a tree barrier management strategy generation module 4.

[0146] The data acquisition module 1 is used to divide the target transmission line into sections and acquire sparse sample data of each section of the transmission line. The sparse sample data includes tree data, terrain data and environmental data.

[0147] The improved generative adversarial network module 2 is used to construct conditional variables based on the sparse sample data, construct real sample data based on the tree data, and input the conditional variables and the real sample data into the trained improved generative adversarial network to obtain generated sample data; the generator loss function of the improved generative adversarial network includes at least a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by the tree age, a tree tilt angle constraint function, and / or a tree growth rate constraint function;

[0148] The Monte Carlo simulation module 3 is used to perform Monte Carlo simulation on the tree sample dataset composed of the real sample data and the generated sample data to obtain the tree-line contact time set corresponding to each section of the transmission line.

[0149] The tree obstacle management strategy generation module 4 is used to generate a tree obstacle management strategy corresponding to each section of the transmission line based on the tree-line contact time set.

[0150] For specific limitations regarding a power transmission line tree obstruction management system, please refer to the above-described limitations regarding a power transmission line tree obstruction management method, which will not be repeated here. Those skilled in the art will recognize that the various modules and steps described in conjunction with the embodiments disclosed in this invention can be implemented in hardware, software, or a combination of both. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0151] like Figure 6 The diagram shows the internal structure of a computer device. An embodiment of the present invention provides a computer device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps described in the embodiment of the power transmission line tree obstruction management method, for example... Figure 1 Steps S1 to S4 as described above.

[0152] Those skilled in the art will understand that the illustrations Figure 6This is merely an example of a computer device and does not constitute a limitation on the computer device. It may include more or fewer components than shown, or combine certain components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0153] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the computer device, connecting various parts of the computer device via various interfaces and lines.

[0154] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0155] If the modules integrated into the computer device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0156] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0157] Accordingly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the steps described in the embodiments of the transmission line tree obstruction management method as described above, for example... Figure 1 Steps S1 to S4 as described above.

[0158] In summary, the embodiments of this application provide a method, system, device, and medium for managing tree obstructions along transmission lines, addressing the technical problem of improving the efficiency, effectiveness, and scientific rigor of such management. The method includes: dividing the target transmission line into segments and acquiring sparse sample data for each segment, including tree data, terrain data, and environmental data; constructing conditional variables based on the sparse sample data and constructing real sample data based on the tree data; inputting the conditional variables and real sample data into a trained improved generative adversarial network (GAN) to obtain generated sample data; the generator loss function of the improved GAN includes at least a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environmental data, a third tree height constraint function determined by tree age, a tree tilt angle constraint function, and / or a tree growth rate constraint function; performing Monte Carlo simulation on the tree sample dataset composed of real sample data and generated sample data to obtain a tree-line contact time set corresponding to each transmission line segment; and generating a tree obstruction management strategy corresponding to each transmission line segment based on the tree-line contact time set. The tree obstacle management method for transmission lines disclosed in this application generates a large amount of generated sample data based on sparse sample data. It does not rely on large-scale sensor deployment or require full historical growth data. It is particularly suitable for transmission lines traversing forest areas, lake areas, and mountainous areas where monitoring is difficult, effectively reducing the management cost and maintenance burden of tree obstacles. By modeling the tree growth process and simulating the behavior of trees in contact with the line, potential contact risks can be predicted in advance. It no longer relies on fixed-period pruning patterns and supports "on-demand pruning." By replacing blind pruning with scientific prediction, it achieves refined management and a shift from passive maintenance to proactive management, reducing the frequency of unnecessary pruning and effectively improving the efficiency of tree obstacle management for transmission lines. At the same time, it provides early warning of high-risk tree obstacles, significantly reducing the occurrence of line tripping and power outages, ensuring the stable operation of the power system, improving power supply reliability, and has good social benefits and promotional value.

[0159] The various embodiments in this specification are described in a progressive manner. For directly identical or similar parts of the embodiments, refer to each other. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. It should be noted that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0160] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the scope of the claims.

Claims

1. A method of managing tree barriers for power line rights-of-way, the method comprising: The method comprises: Divide the target power transmission line surrounding trees according to the tree growth rule to obtain several tree species groups; According to the tree species group and the tree planting year to which the target power transmission line surrounding trees belong, the target power transmission line is divided into several section power transmission lines; Sparse sample data of each section power transmission line is obtained, and the sparse sample data includes tree data, terrain data and environment data; According to the sparse sample data, a condition variable is constructed, and according to the tree data, a real sample data is constructed, and the condition variable and the real sample data are input into a trained improved generative adversarial network to obtain generated sample data; the generator loss function of the improved generative adversarial network at least includes a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environment data, a third tree height constraint function determined by the tree age, a tree inclination angle constraint function and / or a tree growth rate constraint function; Monte Carlo simulation is performed on the tree sample data set composed of the real sample data and the generated sample data to obtain a tree line contact time set corresponding to each section power transmission line; The Monte Carlo simulation on the tree sample data set composed of the real sample data and the generated sample data to obtain a tree line contact time set corresponding to each section power transmission line comprises: According to the real sample data and the generated sample data, a tree sample data set is obtained, and according to the initial height of the tree in the tree sample data set and the tree growth rate, a tree growth function is constructed; According to the tree growth function, a tree line relationship model is constructed, and according to the tree line relationship model, a current tree line distance is obtained; A disturbance factor is introduced into the current tree line distance to obtain a current tree line disturbance distance, and a tree line contact judgment model is constructed according to the current tree line disturbance distance and a pre-set safety radius; According to the tree line contact judgment model, tree line contact judgment is performed on each tree sample data in the tree sample data set to obtain a tree line contact judgment result of each tree sample data in the tree sample data set; According to the tree line contact judgment result, a tree line contact time set corresponding to each section power transmission line is obtained; According to the tree line contact time set, a tree barrier management strategy corresponding to each section power transmission line is generated; According to the tree line contact time set, a tree barrier management strategy corresponding to each section power transmission line is generated, which comprises: Cluster analysis is performed on the tree line contact time set to obtain several cluster centers; According to each cluster center and a pre-set contact time safety margin, a tree barrier management strategy corresponding to each section power transmission line is generated, and the tree barrier management strategy at least includes a tree pruning time sequence.

2. The power line tree obstruction management method of claim 1, wherein, The sparse sample data of each section power transmission line is obtained, which comprises: Collecting tree data of each section power transmission line, the tree data at least including: initial tree height, tree inclination angle, tree growth rate, initial tree age and tree line horizontal distance; collecting terrain data of each section transmission line, the terrain data at least including ground slope; collecting environmental data of each section transmission line, the environmental data at least including soil moisture, light intensity and annual precipitation; constructing sparse sample data according to the tree data, the terrain data and the environmental data.

3. The power line tree obstruction management method of claim 2, wherein, constructing condition variables according to the sparse sample data, and constructing real sample data according to the tree data, comprising: constructing condition variables according to the ground slope, the soil moisture, the light intensity, the annual precipitation and the initial tree age; constructing real sample data according to the initial tree height, the tree inclination angle, the tree growth rate and the tree line horizontal distance.

4. The power line tree obstruction management method of claim 2, wherein, the first tree height constraint function is set to reflect the relationship between the tree height and the ground slope; the second tree height constraint function is set to reflect the relationship between the tree height and the soil moisture; the tree inclination angle constraint function is set to be that the tree inclination angle should be less than the sum of the ground slope and a safety margin; the tree growth rate constraint function is set to be that the tree growth rate is determined by the light intensity and the annual precipitation.

5. The power line tree obstruction management method of claim 1, wherein, the discriminator loss function of the improved generative adversarial network at least includes a gradient penalty term, a real sample score deviation from expectation term and a generated sample score deviation from expectation term; the gradient penalty term is set to reflect the expectation of the gradient of the discriminator on interpolation sample data, the interpolation sample data being set to be interpolated between the real sample data and the generated sample data; the real sample score deviation from expectation term is set to reflect the deviation degree between the first score of the discriminator on the real sample data and the expected score of the real sample data; the generated sample score deviation from expectation term is set to reflect the deviation degree between the second score of the discriminator on the generated sample data and the expected score of the generated sample data.

6. A transmission line tree obstacle management system for implementing the transmission line tree obstacle management method according to any one of claims 1 to 5, characterized by the system comprises a data collection module, an improved generative adversarial network module, a Monte Carlo simulation module and a tree barrier management strategy generation module; the data collection module is used to divide target transmission line surrounding trees according to tree growth rules to obtain several tree species groups; according to the tree species groups and tree planting years of the target transmission line surrounding trees, the target transmission line is divided into several section transmission lines, and sparse sample data of each section transmission line is obtained, the sparse sample data including tree data, terrain data and environmental data; The improved generative adversarial network module is configured to construct a conditional variable according to the sparse sample data, construct real sample data according to the tree data, input the conditional variable and the real sample data into the trained improved generative adversarial network, and obtain generated sample data; the loss function of the generator of the improved generative adversarial network at least includes a first tree height constraint function determined by the terrain data, a second tree height constraint function determined by the environment data, a third tree height constraint function determined by the tree age, a tree inclination angle constraint function, and / or a tree growth rate constraint function; The Monte Carlo simulation module is configured to perform Monte Carlo simulation on a tree sample data set composed of the real sample data and the generated sample data, and obtain a tree-line contact time set corresponding to each section of the transmission line; The tree barrier management strategy generation module is configured to generate a tree barrier management strategy corresponding to each section of the transmission line according to the tree-line contact time set.

7. A computer device, characterized by: The computer device includes a memory, a processor and a transceiver connected through a bus; the memory is configured to store a set of computer program instructions and data, and transmit the stored data to the processor; the processor executes the computer program instructions stored in the memory to perform the transmission line tree barrier management method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer readable storage medium stores a computer program, when the computer program is executed, the transmission line tree barrier management method according to any one of claims 1 to 5 is realized.

Images