Visual multi-dimensional monitoring system for power transmission line based on AI model

Abstract

The invention relates to the technical field of power transmission line monitoring, in particular to a power transmission line visual multi-dimensional monitoring system based on an AI model, which comprises an image acquisition module, an edge calculation module, a deep learning analysis module, an intelligent diagnosis module and a time sequence correlation analysis module. The image acquisition module acquires multispectral data through a visible light camera, an infrared thermal imager and an unmanned aerial vehicle; the edge calculation module adopts an FPGA (Field Programmable Gate Array) chip to realize data preprocessing; the deep learning analysis module comprises a feature fusion sub-module, an anomaly detection sub-module and a three-dimensional reconstruction sub-module which are respectively used for extracting multi-scale features, identifying equipment defects and constructing digital twin bodies; the intelligent diagnosis module integrates the knowledge graph to provide decision support; and the time sequence correlation analysis module is combined with an LSTM-GRU model to realize state prediction. According to the invention, through multi-source data fusion and intelligent analysis, the problems of many blind areas, low identification precision and the like in traditional monitoring are solved, and all-weather and intelligent monitoring, operation and maintenance of the power transmission line are realized.

Description

Technical Field

[0001] This invention relates to the field of power transmission line monitoring technology, and in particular to a multi-dimensional visualization monitoring system for power transmission lines based on an AI model. Background Technology

[0002] With the continuous expansion of power system scale and the sustained increase in voltage levels, the safe operation of transmission lines faces increasingly severe challenges. Traditional manual inspection methods have inherent drawbacks such as low efficiency, high risk, and great dependence on environmental conditions, making them difficult to meet the needs of modern power grids for real-time monitoring of line conditions. In recent years, although image recognition-based line monitoring technology has made some progress, many technical bottlenecks still exist in practical applications.

[0003] In existing technologies, fixed-installation visible light monitoring equipment is often limited by its installation location and viewing angle, making it difficult to comprehensively cover all critical parts of the line, especially in complex terrain areas where there are many monitoring blind spots. Although UAV inspection technology has made up for this deficiency to some extent, its operation mode is mostly periodic inspection along preset routes, lacking intelligent response capabilities for abnormal situations. In addition, the data information collected by a single sensor has limited dimensions. For example, visible light images cannot effectively detect equipment overheating defects, while infrared monitoring is difficult to identify mechanical damage. This data fragmentation severely restricts the accuracy of defect identification.

[0004] At the data analysis level, most existing systems still employ traditional image processing algorithms combined with threshold judgment methods, which are insufficient in identifying subtle defects in complex backgrounds, resulting in high false alarm and false negative rates. Although some advanced systems have introduced deep learning technology, the model architecture is often relatively simple and fails to fully consider the multi-scale characteristics of power transmission equipment, leading to unsatisfactory detection results for small target defects. Furthermore, the decision support functions of these systems are generally weak, lacking the intelligent diagnostic capabilities to combine monitoring results with operational and maintenance knowledge.

[0005] In terms of system architecture, existing solutions mostly adopt a centralized data processing model, where all raw data needs to be transmitted back to the central server. This not only places extremely high demands on communication bandwidth but also causes significant decision-making delays. Although edge computing technology has been applied in other fields, in the scenario of power transmission line monitoring, there is still a lack of mature solutions on how to rationally allocate edge and cloud computing tasks to achieve a balance between real-time performance and analytical depth.

[0006] Furthermore, existing monitoring systems generally suffer from insufficient intelligence, with fixed and rigid data collection strategies that cannot adapt to environmental changes and equipment status, resulting in low resource utilization. In terms of data utilization, they largely remain at the level of simple storage and retrieval, lacking the ability to deeply mine historical data and perform spatiotemporal correlation analysis, making it difficult to achieve true predictive maintenance.

[0007] In summary, current power transmission line visualization monitoring technology still suffers from a series of problems, such as incomplete monitoring, inaccurate identification, untimely response, and unintelligent decision-making. There is an urgent need for a new monitoring system that can integrate multi-source data, achieve intelligent analysis, and support adaptive optimization to meet the needs of modern smart grids for holographic perception and intelligent control of power transmission line status. Summary of the Invention

[0008] To address this, the present invention provides a multi-dimensional visualization monitoring system for transmission lines based on an AI model, which overcomes the significant shortcomings of existing technologies in terms of multi-view data collaboration, computing power optimization, and dynamic early warning mechanisms.

[0009] To achieve the above objectives, on the one hand, the present invention provides a multi-dimensional monitoring system for power transmission lines based on an AI model, including an image acquisition module consisting of a visible light camera, an infrared thermal imager, and a drone inspection device deployed in the power transmission line corridor, used to acquire multispectral images and video stream data of the power transmission equipment;

[0010] The edge computing module includes an image preprocessing submodule, which performs image enhancement, noise reduction and data normalization processing, and integrates an FPGA chip and a 5G communication unit.

[0011] The deep learning analytics module, deployed on a cloud server cluster, includes a feature fusion submodule, an anomaly detection submodule, and a 3D reconstruction submodule; among them,

[0012] The feature fusion submodule uses a hybrid architecture to extract multi-scale spatial features;

[0013] The anomaly detection submodule identifies insulator damage, hardware corrosion, and foreign objects in conductors based on an adaptive threshold dual-stream neural network.

[0014] The three-dimensional reconstruction submodule constructs a digital twin of the power transmission equipment using a motion recovery structure algorithm;

[0015] The intelligent diagnostic module includes a visual human-machine interface submodule, a knowledge graph database submodule, and an AR remote collaboration submodule, which are used to associate historical defect cases, generate maintenance decision suggestions, and support remote collaboration.

[0016] The time-series correlation analysis module uses an LSTM-GRU hybrid model to process continuous monitoring data and performs coupled analysis with the meteorological information system to predict equipment status trends.

[0017] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the image acquisition module further includes an environment adaptive scheduling submodule. This submodule dynamically adjusts the multi-device collaborative acquisition strategy based on real-time meteorological data and equipment operating status, specifically including:

[0018] The exposure parameters of the visible light camera and the start-up threshold of the infrared thermal imager are calculated by using a light intensity sensor and a haze index. When the ambient visibility is lower than the preset value, the infrared device is triggered first and the acquisition of redundant visible light frames is turned off.

[0019] Based on the load current data of the transmission line and the battery endurance of the drone, a multi-objective optimized drone inspection path is generated. In areas with abnormal conductor temperature, the hovering time is extended and visible light and infrared dual-mode data acquisition is initiated simultaneously.

[0020] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the image acquisition module further includes a spatiotemporal correlation triggering submodule. This submodule utilizes historical defect distribution maps and real-time monitoring data to construct spatiotemporal correlation rules, including:

[0021] When the fixed camera detects that the amplitude of the conductor exceeds the safety threshold, it automatically sends high-precision positioning coordinates to the drone and triggers the drone to perform multi-angle image acquisition of the target area using a spiral approximation trajectory.

[0022] During periods of high lightning strike risk, all infrared thermal imagers are forced to operate at the highest sampling rate, and the probability of arc breakdown is predicted based on the temperature rise gradient of the conductor splice tube.

[0023] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the image acquisition module further includes a self-optimizing feedback submodule. This submodule employs a deep reinforcement learning model to score the quality of the acquired data and dynamically update control parameters, including:

[0024] A reward function is established with constraints of image clarity, abnormal area coverage, and data redundancy, and the matching relationship between UAV flight altitude and camera zoom ratio is iteratively optimized.

[0025] By simulating device imaging characteristics under extreme weather conditions using adversarial generative networks, adversarial training samples are generated to improve the robustness of control strategies.

[0026] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the deep learning analysis module further includes a data storage unit for classifying and storing raw image data and processed feature data, wherein:

[0027] When storing raw image data, a three-level index directory is established according to device type, acquisition time, and geographical location; feature data storage uses dual verification of timestamp and device ID to ensure that the mapping relationship with the raw data is traceable.

[0028] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the intelligent diagnostic module is configured to generate multi-level alarm signals based on the anomaly type and confidence level output by the deep learning analysis module, combined with the equipment voltage level.

[0029] For an anomaly where the insulator is damaged and the confidence level is higher than a preset threshold, a red alarm is triggered and automatically pushed to the operation and maintenance terminal;

[0030] If the confidence level of the foreign object detection in the wire is within the preset range, an abnormality will trigger a yellow alarm and be stored in the pending confirmation queue.

[0031] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the intelligent diagnostic module is configured to extract key monitoring indicators and generate structured reports according to a preset cycle: the daily report statistically analyzes the number and type distribution of abnormal events and correlates them with the changing trends of meteorological data;

[0032] The monthly report includes a comparison table of equipment health scores and maintenance work order execution status.

[0033] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the time-series correlation analysis module compares the cloud version number with the local version number to perform differentiated code package download and verification installation.

[0034] The update process retains the core configuration parameters of the old version. If the new version malfunctions, it will automatically roll back to the previous stable version.

[0035] As a preferred technical solution for a multi-dimensional visualization monitoring system for transmission lines based on AI models, the system is equipped with an external data interface that supports data interaction with the power grid SCADA system and the equipment management PMS system: pushing conductor temperature over-limit alarm signals to the SCADA system; and obtaining equipment maintenance records from the PMS system to optimize the knowledge graph database.

[0036] Compared with existing technologies, the advantages of this invention lie in its multi-source data fusion acquisition and intelligent analysis, which significantly improves the comprehensiveness and accuracy of power transmission line monitoring. Through the coordinated operation of visible light, infrared, and UAVs, optimal data acquisition under different environmental conditions is achieved, overcoming the limitations of single-sensor monitoring.

[0037] Furthermore, this invention employs a collaborative architecture of edge computing and cloud-based intelligent analysis, ensuring both real-time data processing and in-depth complex analysis. This architecture effectively balances the allocation of computing resources and improves the overall system operating efficiency.

[0038] Furthermore, the system's early warning mechanism enhances its real-time performance and robustness through closed-loop feedback and adaptive models. Based on feature importance analysis and incremental learning techniques trained on historical data, the fusion weights of multi-source features are dynamically adjusted, enabling the model to adapt to seasonal changes, tree species differences, and environmental fluctuations. In addition, the multi-level risk classification mechanism generates tiered early warning signals through comprehensive quantitative indicators. Combined with the risk area location marking function, this provides maintenance personnel with intuitive decision support, significantly improving the efficiency of problem location and handling.

[0039] Furthermore, the control strategy of this invention enables adaptive and collaborative operation of monitoring equipment. Through environmental perception and intelligent scheduling, the system can automatically adjust the monitoring scheme according to actual operating conditions, significantly improving the targeting of monitoring and resource utilization efficiency.

[0040] Furthermore, the intelligent diagnostic function, supported by knowledge graphs and case libraries, provides maintenance personnel with more accurate fault diagnosis and repair suggestions, reducing the subjectivity and uncertainty of human judgment. The system design boasts excellent scalability and compatibility. Through standardized interface design, it can be easily integrated into existing power monitoring systems, achieving data interconnection and protecting existing user investments. Modular design enables flexible configuration of functions, allowing users to select different function combinations according to actual needs, meeting the monitoring requirements of transmission lines of varying scales and requirements. The overall architecture design fully considers system reliability and stability, ensuring continuous and stable operation of monitoring work through multiple protection mechanisms, providing strong support for the safe operation of transmission lines. Attached Figure Description

[0041] Figure 1 This is a structural block diagram of the AI-based multidimensional monitoring system for power transmission lines, according to an embodiment of the present invention. Detailed Implementation

[0042] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0043] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0044] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0045] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0046] Please see Figure 1 As shown, it is a structural block diagram of the AI-based multidimensional monitoring system for power transmission lines in an embodiment of the present invention. The AI-based multidimensional monitoring system for power transmission lines includes an image acquisition module, which consists of a visible light camera, an infrared thermal imager, and a drone inspection device deployed in the power transmission line corridor, used to acquire multispectral images and video stream data of power transmission equipment.

[0047] The edge computing module includes an image preprocessing submodule, which performs image enhancement, noise reduction and data normalization processing, and integrates an FPGA chip and a 5G communication unit.

[0048] The deep learning analytics module, deployed on a cloud server cluster, includes a feature fusion submodule, an anomaly detection submodule, and a 3D reconstruction submodule; among them,

[0049] The feature fusion submodule uses a hybrid architecture to extract multi-scale spatial features;

[0050] The anomaly detection submodule identifies insulator damage, hardware corrosion, and foreign objects in conductors based on an adaptive threshold dual-stream neural network.

[0051] The three-dimensional reconstruction submodule constructs a digital twin of the power transmission equipment using a motion recovery structure algorithm;

[0052] The intelligent diagnostic module includes a visual human-machine interface submodule, a knowledge graph database submodule, and an AR remote collaboration submodule, which are used to associate historical defect cases, generate maintenance decision suggestions, and support remote collaboration.

[0053] The time-series correlation analysis module uses an LSTM-GRU hybrid model to process continuous monitoring data and couples it with the meteorological information system to predict equipment status trends. Specifically, in the above embodiment, the system's image acquisition module achieves comprehensive monitoring through multi-sensor collaboration. Visible light cameras acquire clear images of the line equipment, identifying mechanical defects such as insulator damage; infrared thermal imagers detect abnormal heat points such as overheated hardware by detecting the surface temperature distribution of the equipment; and unmanned aerial vehicle (UAV) inspection equipment, as a mobile monitoring unit, can supplement inspections of blind spots that are difficult for fixed cameras to cover. This multi-source data acquisition method ensures monitoring without blind spots.

[0054] The edge computing module uses FPGA chips for front-end data processing, primarily handling preprocessing tasks such as image enhancement and normalization. By offloading computational tasks to the edge, closer to the data source, the computational burden on the cloud is reduced while ensuring real-time data transmission. The cloud-based deep learning analysis module is responsible for more complex feature extraction and pattern recognition tasks. Its feature fusion network combines the local feature extraction capabilities of CNNs with the advantages of global relationship modeling, enabling the system to simultaneously capture both detailed device features and overall device status.

[0055] Specifically, the image acquisition module further includes an environment adaptive scheduling submodule, which dynamically adjusts the multi-device collaborative acquisition strategy based on real-time meteorological data and equipment operating status, specifically including:

[0056] The exposure parameters of the visible light camera and the start-up threshold of the infrared thermal imager are calculated by using a light intensity sensor and a haze index. When the ambient visibility is lower than the preset value, the infrared device is triggered first and the acquisition of redundant visible light frames is turned off.

[0057] Based on transmission line load current data and UAV battery endurance, a multi-objective optimized UAV inspection path is generated. In areas with abnormal conductor temperatures, the hovering time is extended, and visible light and infrared dual-modal data acquisition is simultaneously initiated. Understandably, the core of the environmental adaptive scheduling submodule lies in achieving intelligent adjustment of the monitoring strategy. The system dynamically optimizes the collaborative working mode of each sensor by analyzing meteorological conditions and equipment status in real time. For example, under severe weather conditions, the system prioritizes the use of infrared equipment because the quality of visible light imaging will significantly decrease at this time; simultaneously, it adjusts the monitoring focus according to the line load, increasing the monitoring frequency in high-load sections. This adaptive mechanism ensures that effective monitoring data can be obtained under different operating conditions.

[0058] The drone path planning algorithm comprehensively considers factors such as areas with abnormal equipment status and battery life, and uses intelligent scheduling to prioritize limited drone resources to cover the sections that require the most inspection. The system also automatically adjusts the flight route based on the remaining battery power to ensure the drone can return safely.

[0059] To ensure temporal consistency, the image acquisition module further includes a spatiotemporal correlation triggering submodule. This submodule utilizes historical defect distribution maps and real-time monitoring data to construct spatiotemporal correlation rules, including:

[0060] When the fixed camera detects that the amplitude of the conductor exceeds the safety threshold, it automatically sends high-precision positioning coordinates to the drone and triggers the drone to perform multi-angle image acquisition of the target area using a spiral approximation trajectory.

[0061] During periods of high lightning strike risk, all infrared thermal imagers are forced to operate at the highest sampling rate, and the probability of arc breakdown is predicted based on the temperature rise gradient of the conductor splice tube. The spatiotemporal correlation triggering submodule guides real-time monitoring by mining patterns in historical data. The system records and analyzes the spatiotemporal characteristics of line equipment defects, such as the time and location of occurrence, and establishes a probability model for defect occurrence. When real-time monitoring data shows correlation with historical patterns, the system automatically adjusts its monitoring strategy. For example, for sections with historically frequent faults, the monitoring level is increased; under specific seasonal or weather conditions, the detection intensity of related defects is specifically strengthened.

[0062] For dynamic phenomena such as conductor vibration, the system employs a multi-level response mechanism. When abnormal vibration is detected, a fixed camera first makes a preliminary judgment. Once the anomaly is confirmed, a drone is immediately dispatched to the target location for a detailed investigation from multiple angles. This tiered response ensures timely monitoring while avoiding resource waste. Lightning strike risk warning integrates meteorological data and equipment status information to prepare for protective monitoring in advance.

[0063] Specifically, the image acquisition module further includes a self-optimizing feedback submodule, which uses a deep reinforcement learning model to score the quality of the acquired data and dynamically update the control parameters, including:

[0064] A reward function is established with constraints of image clarity, abnormal area coverage, and data redundancy, and the matching relationship between UAV flight altitude and camera zoom ratio is iteratively optimized.

[0065] By simulating device imaging characteristics under extreme weather conditions using generative adversarial networks, adversarial training samples are generated to improve the robustness of the control strategy. This module establishes a closed-loop feedback mechanism to automatically optimize system parameters through continuous evaluation of the acquired data quality. It transforms the performance of each acquisition task into quantifiable metrics, including image sharpness and the integrity of coverage in abnormal areas. By analyzing the correlation between these metrics and acquisition parameters, the system gradually adjusts key parameters such as the UAV's flight altitude and the camera's zoom level, enabling subsequent acquisition tasks to obtain higher-quality data. This self-optimization capability allows the system to adapt to various complex environmental conditions, maintaining good monitoring performance, especially under extreme weather conditions.

[0066] Specifically, the deep learning analysis module further includes a data storage unit for classifying and storing the original image data and the processed feature data, wherein:

[0067] When storing raw image data, a three-level index directory is established based on device type, acquisition time, and geographical location. Feature data storage employs dual verification using timestamps and device IDs to ensure traceability of the mapping relationship with the raw data. This hierarchical and categorized storage strategy enables efficient management of massive amounts of monitoring data. Raw image data is structured and stored according to acquisition device type, timestamp, and geographical location, facilitating rapid retrieval and traceability. Processed feature data uses a lightweight storage format and establishes a bidirectional index relationship with the raw data. This storage architecture ensures both data integrity and improves data utilization efficiency. Simultaneously, the module employs multiple verification mechanisms to ensure data consistency during transmission and storage, providing a reliable data foundation for subsequent analysis.

[0068] Specifically, the intelligent diagnostic module is configured to generate multi-level alarm signals based on the anomaly type and confidence level output by the deep learning analysis module, combined with the device voltage level.

[0069] For an anomaly where the insulator is damaged and the confidence level is higher than a preset threshold, a red alarm is triggered and automatically pushed to the operation and maintenance terminal;

[0070] If the confidence level of the foreign object detection in the wire is within the preset range, an abnormality will trigger a yellow alarm and be stored in the pending confirmation queue.

[0071] Specifically, the intelligent diagnostic module is configured to extract key monitoring indicators and generate structured reports according to a preset cycle: the daily report statistically analyzes the number and type distribution of abnormal events and correlates them with meteorological data change trends; the monthly report integrates a comparison table of equipment health scores and maintenance work order execution status.

[0072] Specifically, the time-series correlation analysis module compares the cloud version number with the local version number to perform differentiated code package download and verification installation:

[0073] The update process retains the core configuration parameters of the old version. If the new version malfunctions, it will automatically roll back to the previous stable version.

[0074] Specifically, the AI-based transmission line visualization multidimensional monitoring system is equipped with an external data interface, which supports data interaction with the power grid SCADA system and the equipment management PMS system: pushing conductor temperature over-limit alarm signals to the SCADA system; and obtaining equipment maintenance records from the PMS system to optimize the knowledge graph database.

[0075] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using dedicated hardware-based apparatus to perform the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0076] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0077] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-dimensional visualization monitoring system for transmission lines based on an AI model, characterized in that, include: The image acquisition module consists of a visible light camera, an infrared thermal imager, and a drone inspection device deployed in the power transmission line corridor, used to acquire multispectral images and video stream data of the power transmission equipment; The edge computing module includes an image preprocessing submodule, which performs image enhancement, noise reduction and data normalization processing, and integrates an FPGA chip and a 5G communication unit. The deep learning analytics module, deployed on a cloud server cluster, includes a feature fusion submodule, an anomaly detection submodule, and a 3D reconstruction submodule; among them, The feature fusion submodule uses a hybrid architecture to extract multi-scale spatial features; The anomaly detection submodule identifies insulator damage, hardware corrosion, and foreign objects in conductors based on an adaptive threshold dual-stream neural network. The three-dimensional reconstruction submodule constructs a digital twin of the power transmission equipment using a motion recovery structure algorithm; The intelligent diagnostic module includes a visual human-machine interface submodule, a knowledge graph database submodule, and an AR remote collaboration submodule, which are used to associate historical defect cases, generate maintenance decision suggestions, and support remote collaboration. The time-series correlation analysis module uses an LSTM-GRU hybrid model to process continuous monitoring data and performs coupled analysis with the meteorological information system to predict equipment status trends.

2. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 1, is characterized in that... The image acquisition module further includes an environment adaptive scheduling submodule, which dynamically adjusts the multi-device collaborative acquisition strategy based on real-time meteorological data and equipment operating status, specifically including: The exposure parameters of the visible light camera and the start-up threshold of the infrared thermal imager are calculated by using a light intensity sensor and a haze index. When the ambient visibility is lower than the preset value, the infrared device is triggered first and the acquisition of redundant visible light frames is turned off. Based on the load current data of the transmission line and the battery endurance of the drone, a multi-objective optimized drone inspection path is generated. In areas with abnormal conductor temperature, the hovering time is extended and visible light and infrared dual-mode data acquisition is initiated simultaneously.

3. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 2, is characterized in that... The image acquisition module further includes a spatiotemporal correlation triggering submodule, which constructs spatiotemporal correlation rules using historical defect distribution maps and real-time monitoring data, including: When the fixed camera detects that the amplitude of the conductor exceeds the safety threshold, it automatically sends high-precision positioning coordinates to the drone and triggers the drone to perform multi-angle image acquisition of the target area using a spiral approximation trajectory. During periods of high lightning strike risk, all infrared thermal imagers are forced to operate at the highest sampling rate, and the probability of arc breakdown is predicted based on the temperature rise gradient of the conductor splice tube.

4. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 3, is characterized in that... The image acquisition module further includes a self-optimizing feedback submodule, which uses a deep reinforcement learning model to score the quality of the acquired data and dynamically update the control parameters, including: A reward function is established with constraints of image clarity, abnormal area coverage, and data redundancy, and the matching relationship between UAV flight altitude and camera zoom ratio is iteratively optimized. By simulating device imaging characteristics under extreme weather conditions using adversarial generative networks, adversarial training samples are generated to improve the robustness of control strategies.

5. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 4, is characterized in that... The deep learning analysis module also includes a data storage unit for classifying and storing the original image data and the processed feature data, wherein: When storing raw image data, a three-level index directory is established according to device type, acquisition time, and geographical location; feature data storage uses dual verification of timestamp and device ID to ensure that the mapping relationship with the raw data is traceable.

6. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 5, is characterized in that... The intelligent diagnostic module is configured to generate multi-level alarm signals based on the anomaly type and confidence level output by the deep learning analysis module, combined with the device voltage level. For an anomaly where the insulator is damaged and the confidence level is higher than a preset threshold, a red alarm is triggered and automatically pushed to the operation and maintenance terminal; If the confidence level of the foreign object detection in the wire is within the preset range, an abnormality will trigger a yellow alarm and be stored in the pending confirmation queue.

7. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 6, is characterized in that... The intelligent diagnostic module is configured to extract key monitoring indicators and generate structured reports according to a preset cycle: the daily report statistically analyzes the number and type distribution of abnormal events and correlates them with the trend of meteorological data changes; The monthly report includes a comparison table of equipment health scores and maintenance work order execution status.

8. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 1, is characterized in that... The time-series correlation analysis module compares the cloud version number with the local version number to perform differentiated code package download and verification installation: The update process retains the core configuration parameters of the old version. If the new version malfunctions, it will automatically roll back to the previous stable version.

9. The AI-based multi-dimensional monitoring system for transmission lines, as described in claim 1, is characterized in that... The AI-based transmission line visualization multidimensional monitoring system is equipped with an external data interface, which supports data interaction with the power grid SCADA system and the equipment management PMS system: pushing conductor temperature over-limit alarm signals to the SCADA system; and obtaining equipment maintenance records from the PMS system to optimize the knowledge graph database.

Images