Power transmission line anomaly detection method and device

Abstract

The invention discloses a power transmission line anomaly detection method and device, and belongs to the field of power transmission line anomaly detection.The method comprises the steps that multi-mode monitoring data of a power transmission line are obtained; performing feature extraction and fusion on the multi-modal monitoring data by adopting a spatial-temporal feature fusion algorithm to generate a spatial-temporal correlation feature matrix; inputting the space-time correlation feature matrix into a trained deep residual adversarial neural network model, and outputting an anomaly probability distribution diagram which comprises the probability condition of anomaly at each position on the power transmission line; and analyzing the anomaly probability distribution diagram through a dynamic threshold segmentation algorithm, and determining the type and the position of the anomaly occurring in the power transmission line. According to the method, the multi-modal monitoring data and the spatial-temporal feature fusion algorithm are combined, the complementarity of different types of data is fully utilized, and the accuracy and reliability of anomaly detection are effectively improved; the dynamic threshold segmentation algorithm considers line operation conditions and environmental parameters, and reduces false alarms and missing alarms.

Description

Technical Field

[0001] This invention relates to a method and apparatus for detecting anomalies in power transmission lines, belonging to the field of power transmission line anomaly detection. Background Technology

[0002] Power lines are exposed to complex and ever-changing natural environments and operating conditions for a long time, making them susceptible to various factors such as equipment aging, lightning strikes, icing, and external damage, which can lead to various abnormalities and malfunctions.

[0003] Traditional methods for detecting anomalies in power transmission lines mainly rely on manual inspections, which are not only inefficient and labor-intensive, but also suffer from long inspection cycles, high missed detection rates, and difficulty in identifying potential hazards. With the development of sensor technology, information technology, and artificial intelligence, detection methods based on sensor monitoring and data analysis are gradually being applied. However, existing detection methods often have the following shortcomings: the data information acquired by a single sensor is limited and cannot comprehensively reflect the line status; multi-source data fusion processing technology is not perfect, making it difficult to effectively extract data features; in summary, existing technologies still suffer from poor accuracy and real-time performance in power transmission line anomaly detection, and cannot meet the requirements for high-precision detection. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for detecting anomalies in power transmission lines, thereby solving the problems of poor accuracy and real-time performance in the existing technology for detecting anomalies in power transmission lines.

[0005] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, the present invention provides a method for detecting anomalies in power transmission lines, comprising: Acquire multimodal monitoring data of transmission lines; A spatiotemporal feature fusion algorithm is used to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The spatiotemporal correlation feature matrix is ​​input into the trained deep residual adversarial neural network model, and an anomaly probability distribution map is output, which includes the probability of anomalies occurring at each location on the transmission line. The abnormal probability distribution map is analyzed by a dynamic threshold segmentation algorithm to determine the type and location of abnormalities in the transmission line.

[0006] Furthermore, the multimodal monitoring data includes various monitoring data from multiple monitoring nodes in the transmission line; The monitoring data includes infrared thermal imaging data, ultraviolet discharge data, visible light image data, and vibration spectrum data.

[0007] Furthermore, the step of using a spatiotemporal feature fusion algorithm to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix includes: A three-dimensional convolutional neural network was used to extract spatial-temperature features from infrared thermal imaging data. Wavelet packet transform was used to extract the time-frequency features of vibration spectrum data; The spatial-temperature features and the time-frequency features are weighted and fused using an attention mechanism to generate a spatiotemporal correlation feature matrix.

[0008] Furthermore, the deep residual adversarial neural network model is trained using the following method: Construct a sample library, which contains monitoring data categorized as normal and multiple categories of abnormal, with no fewer than 5,000 samples for each type of monitoring data; The samples in the sample library are input into the generative adversarial network for adversarial training to obtain multiple sets of high-quality samples. The generative adversarial network includes a generator and a discriminator. The generator is a deep residual network with skip connections, and the discriminator is a multi-scale convolutional network. A large-scale power equipment dataset is obtained, and the deep residual adversarial neural network model is pre-trained using the data in the large-scale power equipment dataset. Then, the deep residual adversarial neural network model is trained using multiple sets of high-quality samples to obtain a trained deep residual adversarial neural network model.

[0009] Furthermore, the step of analyzing the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies in the transmission line includes: Obtain the constructed three-dimensional dynamic threshold matrix; After obtaining the current operating conditions, environmental parameters and equipment type of the transmission line, the corresponding dynamic threshold is retrieved from the three-dimensional dynamic threshold matrix. If the anomaly probability value at a certain location in the anomaly probability distribution map exceeds the dynamic threshold and continues for more than three sampling periods, then a valid anomaly has occurred at that location. Based on multimodal monitoring data and spatiotemporal correlation feature matrix, the type of the valid anomaly is determined.

[0010] Furthermore, the operating conditions include current load and voltage level, the environmental parameters include temperature, humidity and wind speed, and the equipment types include conductors, insulators and towers.

[0011] Furthermore, after obtaining the current operating conditions, environmental parameters, and equipment type of the transmission line, the corresponding dynamic threshold is retrieved from the three-dimensional dynamic threshold matrix, including: After obtaining the current operating conditions, environmental parameters, and equipment type of the transmission line, the K-nearest neighbor algorithm is used to match the most similar set of operating conditions, environmental parameters, and equipment type in the three-dimensional dynamic threshold matrix, and the corresponding dynamic threshold is called as the dynamic threshold for subsequent judgment.

[0012] Furthermore, it also includes a step of periodically calibrating the three-dimensional dynamic threshold matrix: If the false alarm rate exceeds 5% under a certain type of operating condition, the dynamic threshold corresponding to that type of operating condition will be lowered by 0.05. If the false negative rate exceeds 3% under a certain type of operating condition, the dynamic threshold corresponding to that type of operating condition will be increased by 0.03.

[0013] Furthermore, after retrieving the corresponding dynamic threshold from the three-dimensional dynamic threshold matrix, the method further includes: If the anomaly probability value at a certain location in the anomaly probability distribution map does not exceed the dynamic threshold, but the local anomaly factor is greater than 1.5 and lasts for more than three sampling periods, then a potential micro-anomaly has appeared at that location, an early warning instruction is issued, and the location is marked as a key monitoring location. The local anomaly factor is obtained by the following method: the anomaly probability distribution map is analyzed locally using a sliding window statistical method to obtain the local anomaly factor.

[0014] Secondly, the present invention provides a transmission line anomaly detection device, comprising: The data acquisition module is configured to acquire multimodal monitoring data of transmission lines; The feature extraction and fusion module is configured to: use a spatiotemporal feature fusion algorithm to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The anomaly probability calculation module is configured to: input the spatiotemporal correlation feature matrix into the trained deep residual adversarial neural network model and output an anomaly probability distribution map, wherein the anomaly probability distribution map includes the probability of anomalies occurring at each location on the transmission line; The anomaly detection module is configured to analyze the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies occurring in the transmission line.

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are: This invention provides a method and apparatus for detecting anomalies in transmission lines. It utilizes multimodal monitoring data for anomaly detection and employs a spatiotemporal feature fusion algorithm for processing. This fully leverages the complementarity of different data types, comprehensively and accurately reflecting the operating status of transmission lines and effectively improving the accuracy and reliability of anomaly detection. A deep residual adversarial neural network model combined with a transfer learning strategy enhances the model's generalization ability and training efficiency. A dynamic threshold segmentation algorithm considers line operating conditions and environmental parameters, enabling more precise anomaly identification and reducing false alarms and missed alarms. Attached Figure Description

[0016] Figure 1 This is a flowchart of a transmission line anomaly detection method corresponding to Example 1. Detailed Implementation

[0017] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solution of the present invention more clearly, and should not be used to limit the scope of protection of the present invention.

[0018] Example 1

[0019] like Figure 1 As shown, this embodiment provides a method for detecting anomalies in transmission lines, including: Acquire multimodal monitoring data of transmission lines; A spatiotemporal feature fusion algorithm is used to extract and fuse features from multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The spatiotemporal correlation feature matrix is ​​input into the trained deep residual adversarial neural network model, and the output is an anomaly probability distribution map, which includes the probability of anomalies occurring at various locations on the transmission line. By analyzing the anomaly probability distribution map using a dynamic threshold segmentation algorithm, the type and location of anomalies in the transmission line can be determined.

[0020] This invention implements anomaly detection using multimodal monitoring data and processes it with a spatiotemporal feature fusion algorithm. It fully utilizes the complementarity of different types of data, which can comprehensively and accurately reflect the operating status of transmission lines, effectively improving the accuracy and reliability of anomaly detection. The deep residual adversarial neural network model combined with the transfer learning strategy improves the model's generalization ability and training efficiency. The dynamic threshold segmentation algorithm considers the line operating conditions and environmental parameters, which can more accurately determine anomalies and reduce false alarms and missed alarms.

[0021] Example 2

[0022] This embodiment provides a method for detecting anomalies in power transmission lines, specifically including the following steps: S1. Acquire multimodal monitoring data of transmission lines.

[0023] In practical applications, monitoring nodes are strategically placed along the transmission line at intervals of 200-300 meters. Each monitoring node is equipped with at least three different types of sensors, including infrared thermal imaging sensors, ultraviolet discharge sensors, visible light image sensors, and vibration sensors. Data is acquired between the sensors through an adaptive synchronous sampling protocol. When the line load is low, the sampling frequency is appropriately reduced to decrease the amount of data processing; when the line load increases, the sampling frequency is automatically increased to ensure that subtle changes in the line's operating status can be captured, thereby obtaining comprehensive and accurate multimodal monitoring data. S2. A spatiotemporal feature fusion algorithm is used to extract and fuse features from multimodal monitoring data to generate a spatiotemporal correlation feature matrix.

[0024] For the acquired infrared thermal imaging data, a three-dimensional convolutional neural network is used to process it. Convolution operations are employed to extract spatial-temperature features from the images, capturing the spatial information and temperature variation characteristics of the line equipment's temperature distribution. For vibration spectrum data, wavelet packet transform technology is used to decompose the signal into different frequency bands, thereby extracting time-frequency features reflecting the line's vibration characteristics. Then, an attention mechanism is used to weight and fuse the feature vectors extracted from different modes, assigning different weights to the importance of anomaly detection based on the features of each mode, forming a unified spatiotemporal correlation feature matrix containing spatiotemporal information.

[0025] S3. Input the spatiotemporal correlation feature matrix into the trained deep residual adversarial neural network model and output an anomaly probability distribution map, which includes the probability of anomalies occurring at various locations on the transmission line.

[0026] In this embodiment, the training of the deep residual adversarial neural network model includes: First, constructing a rich sample library covering the normal operation status of transmission lines and data samples of 12 typical abnormal states such as conductor strand breakage, insulator damage, and tower tilting, with no less than 5000 samples in each class. A generative adversarial network (GAN) framework is adopted. The generator uses a deep residual network structure with skip connections, which effectively alleviates the gradient vanishing problem during network training and enhances the network's learning ability. The discriminator uses a multi-scale convolutional network to better identify data features. Samples from the sample library are input into the GAN for adversarial training to obtain multiple sets of high-quality samples. During training, a transfer learning strategy is introduced. The model is first pre-trained on a large-scale power equipment dataset to learn the general feature representation of power equipment data. Then, it is fine-tuned on the target transmission line dataset (including multiple sets of high-quality samples) to adapt the model to the specific task of transmission line anomaly detection.

[0027] S4. Analyze the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies in the transmission line.

[0028] A three-dimensional dynamic threshold matrix is ​​pre-constructed based on the transmission line's past operating conditions (such as current load, voltage level, etc.), environmental parameters (such as temperature, humidity, wind speed, etc.), and equipment type (conductors, insulators, towers, etc.). For example: For abnormal overheating of conductors, when the current load is 80% of the rated value and the ambient temperature is 35℃, the corresponding dynamic threshold is 0.65. Under the same load, if the ambient temperature drops to 10℃, the dynamic threshold will be automatically adjusted to 0.58 to adapt to normal temperature fluctuations in low-temperature environments.

[0029] The system collects real-time parameters (current load, temperature, humidity, etc.) every 100ms. It then uses the K-Nearest Neighbor (KNN) algorithm to match the most similar operating condition in a three-dimensional dynamic threshold matrix and calls the corresponding dynamic threshold as the current judgment criterion. For example: When the current load is 75% and the temperature is 30℃, the system automatically retrieves the average dynamic threshold value (e.g., 0.62) in the "70-80% load + 28-32℃" range from the three-dimensional dynamic threshold matrix and uses it as the current anomaly judgment threshold.

[0030] The three-dimensional dynamic threshold matrix is ​​calibrated monthly based on historical data: if the false alarm rate exceeds 5% under a certain operating condition, the dynamic threshold for that operating condition is automatically lowered by 0.05; if the false alarm rate exceeds 3%, it is raised by 0.03 to ensure the adaptability of the dynamic threshold.

[0031] Obtain the constructed three-dimensional dynamic threshold matrix; After obtaining the current operating conditions, environmental parameters and equipment type of the transmission line, the corresponding dynamic threshold is retrieved from the three-dimensional dynamic threshold matrix. If the anomaly probability value at a certain location in the anomaly probability distribution map exceeds the dynamic threshold and continues for more than three sampling periods, then a valid anomaly has occurred at that location. Based on multimodal monitoring data and spatiotemporal correlation feature matrix, the types of valid anomalies are determined.

[0032] After retrieving the corresponding dynamic threshold from the three-dimensional dynamic threshold matrix, the following is also included: If the anomaly probability value at a certain location in the anomaly probability distribution map does not exceed the dynamic threshold, but the local anomaly factor is greater than 1.5 and lasts for more than three sampling periods, then a potential micro-anomaly has appeared at that location, an early warning instruction is issued, and the location is marked as a key monitoring location. The local anomaly factors are obtained by performing local analysis on the anomaly probability distribution map using the sliding window statistical method.

[0033] The multi-source data acquisition module continuously acquires multimodal monitoring data of the transmission line through a distributed sensor array and transmits the data to edge computing nodes. The FPGA (Field Programmable Gate Array) chip on the edge computing node performs real-time preprocessing on the raw sensor data, such as data filtering and format conversion. The GPU (Graphics Processing Unit) coprocessor uses a lightweight neural network inference engine to perform preliminary analysis on the preprocessed data and extract some features. The edge computing node transmits the processed data and features to the cloud analysis platform through a communication module with time-sensitive networking capabilities.

[0034] After receiving the data, the cloud-based analytics platform uses a deployed deep residual adversarial neural network model for further analysis, and combines this with an expert knowledge base to make a comprehensive judgment on anomalies. The digital twin module constructs a three-dimensional virtual model of the transmission line and updates the model's status in real time based on real-time monitoring data, achieving a precise mapping between the physical line and the virtual model. The predictive maintenance module uses data analysis algorithms to predict the remaining service life of equipment based on historical anomaly data and the current line status. The adaptive learning module optimizes the detection model based on newly discovered anomaly samples, continuously improving the model's detection accuracy.

[0035] The human-machine interface terminal receives anomaly detection results from the cloud-based analysis platform and displays them to operations and maintenance personnel in a visual manner, such as marking anomaly locations on a map and displaying anomaly types and probabilities in charts. It also provides corresponding handling suggestions based on an expert knowledge base. When an anomaly is detected, the drone collaborative verification subsystem automatically dispatches the nearest drone to the anomaly location for close-range verification to obtain more detailed anomaly information. The blockchain evidence storage module records data throughout the entire anomaly detection process, ensuring data authenticity and immutability. The intelligent alarm module automatically triggers corresponding emergency response procedures based on the anomaly level, such as notifying relevant operations and maintenance personnel and activating emergency plans.

[0036] By adopting an architecture that combines edge computing and cloud analytics, distributed data processing and centralized analysis are achieved, improving the system's real-time performance and processing capabilities. Functional modules such as digital twins and predictive maintenance provide strong support for the intelligent operation and maintenance of transmission lines.

[0037] The inclusion of a drone-based collaborative verification subsystem, a blockchain-based evidence storage module, and an intelligent alarm module further enhances the system's intelligence level and operational efficiency, ensuring the scientific, standardized, and timely detection and handling of power transmission line anomalies. Example 3

[0038] This embodiment provides a power transmission line anomaly detection device, including: The data acquisition module is configured to acquire multimodal monitoring data of transmission lines; The feature extraction and fusion module is configured to: use a spatiotemporal feature fusion algorithm to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The anomaly probability calculation module is configured to: input the spatiotemporal correlation feature matrix into the trained deep residual adversarial neural network model and output an anomaly probability distribution map, wherein the anomaly probability distribution map includes the probability of anomalies occurring at each location on the transmission line; The anomaly detection module is configured to analyze the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies occurring in the transmission line.

[0039] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0040] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0041] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0042] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0043] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for detecting anomalies in transmission lines, characterized in that, include: Acquire multimodal monitoring data of transmission lines; A spatiotemporal feature fusion algorithm is used to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The spatiotemporal correlation feature matrix is ​​input into the trained deep residual adversarial neural network model, and an anomaly probability distribution map is output, which includes the probability of anomalies occurring at each location on the transmission line. The abnormal probability distribution map is analyzed by a dynamic threshold segmentation algorithm to determine the type and location of abnormalities in the transmission line.

2. The method for detecting anomalies in transmission lines according to claim 1, characterized in that, The multimodal monitoring data includes various monitoring data from multiple monitoring nodes in the transmission line; The monitoring data includes infrared thermal imaging data, ultraviolet discharge data, visible light image data, and vibration spectrum data.

3. The method for detecting abnormalities in transmission lines according to claim 1, characterized in that, The process of using a spatiotemporal feature fusion algorithm to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix includes: A three-dimensional convolutional neural network was used to extract spatial-temperature features from infrared thermal imaging data. Wavelet packet transform was used to extract the time-frequency features of vibration spectrum data; The spatial-temperature features and the time-frequency features are weighted and fused using an attention mechanism to generate a spatiotemporal correlation feature matrix.

4. The method for detecting abnormalities in transmission lines according to claim 1, characterized in that, The deep residual adversarial neural network model is trained using the following method: Construct a sample library, which contains monitoring data categorized as normal and multiple categories of abnormal, with no fewer than 5,000 samples for each type of monitoring data; The samples in the sample library are input into the generative adversarial network for adversarial training to obtain multiple sets of high-quality samples. The generative adversarial network includes a generator and a discriminator. The generator is a deep residual network with skip connections, and the discriminator is a multi-scale convolutional network. A large-scale power equipment dataset is obtained, and the deep residual adversarial neural network model is pre-trained using the data in the large-scale power equipment dataset. Then, the deep residual adversarial neural network model is trained using multiple sets of high-quality samples to obtain a trained deep residual adversarial neural network model.

5. The method for detecting anomalies in transmission lines according to claim 1, characterized in that, The step of analyzing the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies in the transmission line includes: Obtain the constructed three-dimensional dynamic threshold matrix; After obtaining the current operating conditions, environmental parameters and equipment type of the transmission line, the corresponding dynamic threshold is retrieved from the three-dimensional dynamic threshold matrix. If the anomaly probability value at a certain location in the anomaly probability distribution map exceeds the dynamic threshold and continues for more than three sampling periods, then a valid anomaly has occurred at that location. Based on multimodal monitoring data and spatiotemporal correlation feature matrix, the type of the valid anomaly is determined.

6. The method for detecting anomalies in transmission lines according to claim 5, characterized in that, The operating conditions include current load and voltage level; the environmental parameters include temperature, humidity and wind speed; and the equipment types include conductors, insulators and towers.

7. The method for detecting abnormalities in transmission lines according to claim 5, characterized in that, After obtaining the current operating conditions, environmental parameters, and equipment type of the transmission line, the corresponding dynamic threshold is retrieved from the three-dimensional dynamic threshold matrix, including: After obtaining the current operating conditions, environmental parameters, and equipment type of the transmission line, the K-nearest neighbor algorithm is used to match the most similar set of operating conditions, environmental parameters, and equipment type in the three-dimensional dynamic threshold matrix, and the corresponding dynamic threshold is called as the dynamic threshold for subsequent judgment.

8. The method for detecting abnormalities in transmission lines according to claim 5, characterized in that, It also includes a step of periodically calibrating the three-dimensional dynamic threshold matrix: If the false alarm rate exceeds 5% under a certain type of operating condition, the dynamic threshold corresponding to that type of operating condition will be lowered by 0.

05. If the false negative rate exceeds 3% under a certain type of operating condition, the dynamic threshold corresponding to that type of operating condition will be increased by 0.

03.

9. The method for detecting abnormalities in transmission lines according to claim 5, characterized in that, After retrieving the corresponding dynamic threshold from the three-dimensional dynamic threshold matrix, the method further includes: If the anomaly probability value at a certain location in the anomaly probability distribution map does not exceed the dynamic threshold, but the local anomaly factor is greater than 1.5 and lasts for more than three sampling periods, then a potential micro-anomaly has appeared at that location, an early warning instruction is issued, and the location is marked as a key monitoring location. The local anomaly factor is obtained by the following method: the anomaly probability distribution map is analyzed locally using a sliding window statistical method to obtain the local anomaly factor.

10. A transmission line anomaly detection device, characterized in that, include: The data acquisition module is configured to acquire multimodal monitoring data of transmission lines; The feature extraction and fusion module is configured to: use a spatiotemporal feature fusion algorithm to extract and fuse features from the multimodal monitoring data to generate a spatiotemporal correlation feature matrix; The anomaly probability calculation module is configured to: input the spatiotemporal correlation feature matrix into the trained deep residual adversarial neural network model and output an anomaly probability distribution map, wherein the anomaly probability distribution map includes the probability of anomalies occurring at each location on the transmission line; The anomaly detection module is configured to analyze the anomaly probability distribution map using a dynamic threshold segmentation algorithm to determine the type and location of anomalies occurring in the transmission line.

Images