On-line monitoring system for conductor sag of power transmission line

Abstract

The invention discloses a power transmission line conductor sag on-line monitoring system, and relates to the technical field of power transmission line sag monitoring. The positioning calibration module is used for sending a power transmission line locking instruction to enable the unmanned aerial vehicle to perform positioning calibration and obtain a video stream; the curve fitting module is used for performing feature extraction on the image of the video stream to obtain line feature fitting to obtain a space curve equation; the sag calculation module is used for calculating according to the space curve equation to obtain a first sag numerical value, acquiring attribute characteristics of the power transmission line in real time and calculating to obtain a second sag numerical value; and the fault early warning module is used for carrying out fault early warning if the absolute difference value between the first sag numerical value and the second sag numerical value is greater than a risk threshold value. The power transmission line is monitored through real-time video streaming, changes are accurately captured in combination with feature extraction and curve fitting, sag is calculated in real time, and the state is evaluated. The system performs early warning when the sag is abnormal, reduces false alarm based on accurate data, improves early warning accuracy, ensures stable operation of a power grid, and timely discovers potential fault risks.

Description

Technical Field

[0001] This invention belongs to the field of transmission line sag monitoring technology, specifically relating to an online monitoring system for transmission line conductor sag. Background Technology

[0002] Conductor sag, as a core parameter in the design and operation of transmission lines, directly relates to the safety and reliability of the power grid, affecting the mechanical, electrical, and environmental adaptability of the lines. A reasonable conductor sag design ensures stable operation of the lines under various conditions, avoids line faults caused by excessive or insufficient sag, and thus guarantees continuous power supply and stable operation of the power grid.

[0003] Patent No. CN110307825A discloses an online monitoring system for the sag of overhead transmission lines. The system includes tilt angle measuring spheres at both ends of the transmission line suspension. Since the measurement of the suspension point tilt angle is often subject to errors due to mechanical structural factors and sensor issues, a ground distance measuring sphere is also installed at the lowest point of the transmission line sag. By combining this with direct measurement of the sag to the ground, the calculated sag results can be corrected, effectively correcting errors often caused by mechanical structural factors and sensor issues in the measurement of the suspension point tilt angle, thus greatly improving the measurement accuracy of the sag. Furthermore, by directly measuring the sag to the ground, the actual height of the conductor above the ground can be obtained, providing more effective information support for line maintenance in the event of unforeseen events below the conductor, such as illegal construction, tree growth, or construction machinery.

[0004] The existing method for determining conductor sag is through manual measurement. However, this method suffers from low measurement efficiency, poor real-time performance, and insufficient accuracy due to the low precision of the monitoring equipment. Summary of the Invention

[0005] The purpose of this invention is to solve the problems of low accuracy of monitoring equipment, which leads to low measurement efficiency, poor real-time performance and insufficient precision, and to propose an online monitoring system for conductor sag of transmission lines.

[0006] In a first aspect of the invention, an online monitoring system for conductor sag of a transmission line is first proposed, the system comprising: a positioning calibration module, a curve fitting module, a sag calculation module, and a fault early warning module.

[0007] The positioning calibration module is used to send a power line locking command to enable the UAV to position and calibrate, and to acquire a video stream of the power line in the work area;

[0008] The curve fitting module is used to extract features from the video stream image to obtain line features, and to fit the line features to obtain a spatial curve equation.

[0009] The sag calculation module is used to calculate the first sag value according to the spatial curve equation, acquire the attribute characteristics of the transmission line in real time, and calculate the second sag value according to the attribute characteristics.

[0010] The fault warning module is used to issue a transmission line fault warning if the absolute difference between the first sag value and the second sag value is greater than a risk threshold.

[0011] Optionally, the positioning calibration module further includes: a power tower identification module, a first control module, a second control module, and a mapping calibration module.

[0012] The power tower identification module is used to acquire image data of power transmission lines in real time to obtain a first target image set, and to perform power tower identification and detection on the first target images in the target image set; the target image set includes multiple images.

[0013] The first control module is used to perform a uniform speed operation to move away from the power transmission line according to the drone's positioning if the first target image does not identify the power tower, until the power tower is identified in the first target image;

[0014] The second control module is used to, if the first target image detects a power tower, perform a reverse horizontal rotation according to the position of the power tower in the first target image until a set of power towers appears in the first target image; a set of power towers includes two power towers.

[0015] The composition calibration module is used to perform composition calibration based on the position of the group of power towers in the first target image, and then capture a video stream after calibration.

[0016] Optionally, the first control module includes: an image acquisition module, a feature fusion module, a coordinate determination module, a UAV positioning module, and an altitude determination module.

[0017] The image acquisition module is used to acquire measured images and satellite images captured by the UAV, and to randomly cut the measured images into sub-image sets according to a preset size;

[0018] The feature fusion module is used to extract feature maps from measured images, satellite images, and sub-image sets through the backbone network to obtain a feature map set, and to fuse the feature maps in the feature map set to obtain the final feature map;

[0019] The coordinate determination module is used to generate a target heat map based on the final feature map, and to determine the UAV coordinates based on the heat distribution results in the target heat map.

[0020] The UAV positioning module is used to interpolate the feature maps in the feature map set to obtain a target satellite image, and to determine the position of the UAV in the target satellite image based on the UAV coordinates;

[0021] The altitude determination module is used to determine the drone altitude based on the position of the measured image in the satellite image and the scaling ratio, and based on the scaling ratio.

[0022] Optionally, the curve fitting module includes: a historical data acquisition module, a data training module, a model update module, and a feature extraction module.

[0023] The historical data acquisition module is used to acquire a set of historical spacer images and a set of historical transmission line images, and to use the set of historical spacer images and the set of historical transmission line images as input to a convolutional neural network model for training; the historical transmission line images include historical spacer images;

[0024] The data training module is used to extract historical spacer images and historical transmission line images through a feature extraction network to obtain feature maps, determine candidate regions in the feature maps through a region candidate network to obtain candidate region feature maps, and input the candidate region feature maps into the head network through the ROI Align layer for evaluation to obtain the final candidate regions.

[0025] The model update module is used to use the final candidate region as the model update parameter, and update each layer in the convolutional neural network model according to the model update parameter to obtain the feature extraction model.

[0026] The feature extraction module is used to divide the video stream according to a preset period to obtain a second target image, and input the second target image into the feature extraction model for training to obtain line features; the line features include multiple feature points.

[0027] Optionally, the curve fitting module is further configured to determine the center coordinates of each feature point based on the feature points in the line features, fit the center coordinates of each feature point to obtain the spatial curve equation through a spatial fitting algorithm, and calculate the first sag value.

[0028] Optionally, the sag calculation module includes: a multi-parameter curve equation fitting module and a second sag numerical calculation module.

[0029] The multi-parameter curve equation fitting module is used to acquire the attribute characteristics of the transmission line in real time, and to obtain the multi-parameter curve equation by fitting the curve equation based on the attribute characteristics.

[0030] The second sag value calculation module is used to calculate the second sag value based on the multi-parameter curve equation;

[0031]

[0032] Where f0 is the sag value, g is the specific load, L is the horizontal distance between the two suspension points, k0 is the tension of the conductor at the suspension point, α is the coefficient of thermal expansion of the conductor, and ΔT is the change in temperature.

[0033] Optionally, the transmission line fault early warning system includes: a risk level assessment module, a first early warning module, and a second early warning module.

[0034] The risk level assessment module is used to upload the fault location of the transmission line to the system and perform a risk level assessment if the absolute difference between the first sag value and the second sag value is greater than the risk threshold.

[0035] The first early warning module is used to increase the monitoring frequency of the transmission line within a preset time period if the risk threshold is less than the absolute difference and less than the high risk threshold.

[0036] The second early warning module is used to issue a fault alarm and upload the monitoring data of the transmission line to the cloud if the high-risk threshold is less than the absolute difference.

[0037] The beneficial effects of this invention are as follows: The online conductor sag monitoring system of this invention includes a positioning calibration module for receiving transmission line locking commands and performing UAV positioning calibration to acquire video streams of the transmission lines within the work area; a curve fitting module for extracting features from the transmission lines in the video stream to obtain line features, and fitting a spatial curve equation based on these features; a sag calculation module for calculating a first sag value based on the spatial curve equation, acquiring the attribute features of the transmission lines in real time, and calculating a second sag value based on these attribute features; and a fault early warning module for issuing a transmission line fault early warning if the absolute difference between the first and second sag values ​​exceeds a risk threshold. By monitoring the transmission lines in real time through video stream, combined with feature extraction and spatial curve equation fitting, the system accurately captures line changes, calculates sag values ​​in real time, and assesses the sag status. When the absolute difference between the sag values ​​exceeds the risk threshold, the system immediately issues an early warning, promptly identifying potential fault risks. Based on accurate data analysis, it reduces false alarms and missed alarms, improves the accuracy and reliability of early warnings, and ensures the stable operation of the power grid. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the online conductor sag monitoring system of the present invention;

[0039] Figure 2 This is a schematic diagram of the principle of the online monitoring system for conductor sag of the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and B can represent: A alone, A and B simultaneously, and B alone. Furthermore, descriptions involving "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" can explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0041] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] This invention provides an online monitoring system for conductor sag in power transmission lines. See also... Figure 1 , Figure 1 A flowchart illustrating an online monitoring system for conductor sag in a power transmission line, provided as an embodiment of the present invention. The system includes: a positioning and calibration module, a curve fitting module, a sag calculation module, and a fault early warning module.

[0043] The positioning and calibration module is used to send power line locking commands to enable the UAV to position and calibrate, and to acquire video streams of power lines within the work area;

[0044] The curve fitting module is used to extract features from the video stream to obtain line features, and then fit the line features to obtain the spatial curve equation.

[0045] The sag calculation module is used to calculate the first sag value based on the spatial curve equation, and to obtain the attribute characteristics of the transmission line in real time and calculate the second sag value based on the attribute characteristics.

[0046] The fault warning module is used to issue a fault warning for the transmission line if the absolute difference between the first sag value and the second sag value is greater than the risk threshold.

[0047] The online monitoring system for conductor sag of a transmission line provided by this invention monitors the transmission line through real-time video streams. By combining feature extraction and spatial curve equation fitting, it accurately captures line changes, calculates sag values ​​in real time, and assesses the sag status. When the absolute difference in sag values ​​exceeds a risk threshold, the system immediately issues an early warning, promptly identifying potential fault risks. Based on accurate data analysis, it reduces false alarms and missed alarms, improves the accuracy and reliability of early warnings, and ensures the stable operation of the power grid.

[0048] In one implementation, see [link to implementation details]. Figure 2 , Figure 2 This is a schematic diagram of an online monitoring system for conductor sag of a power transmission line provided in an embodiment of the present invention. A represents the power tower, B represents the power transmission line, and C represents the sag value. By receiving a power transmission line locking command and performing UAV positioning calibration, the system can quickly and accurately acquire video streams of the power transmission line within the work area, shortening the monitoring response time. The natural sag of the power transmission line due to gravity forms a curve. By identifying and sampling points on this curve, a spatial curve equation is obtained. The connection between the power transmission line and the power tower is used to obtain a line segment. The distance from the lowest point on the spatial curve equation to the line segment is taken as the sag value.

[0049] In one implementation, the system calculates a first sag value (score) based on the spatial curve equation, acquires the attribute characteristics of the transmission line in real time, and calculates a second sag value (score) based on the attribute characteristics. Comparing the first and second sag values ​​helps to gain a more comprehensive understanding of the sag state of the transmission line, reducing errors that may be caused by a single calculation method. By comparing the absolute difference between the two sag values, the system can more accurately determine whether there is an anomaly in the transmission line, thus improving the accuracy of fault early warning.

[0050] In one implementation, when the absolute difference between the first sag value and the second sag value exceeds a preset risk threshold, the system immediately issues a transmission line fault warning. This warning mechanism can detect potential risks in time before a fault occurs, providing maintenance personnel with ample time for inspection and maintenance. Because the system's warning is based on precise data analysis and calculation, it greatly reduces false alarms and missed alarms, improving the accuracy and reliability of the warning.

[0051] In one embodiment, the positioning calibration module further includes: a power tower identification module, a first control module, a second control module, and a mapping calibration module.

[0052] The power tower identification module is used to acquire image data of power transmission lines in real time to obtain a first target image set, and to perform power tower identification and detection on the first target image in the target image set; the target image set includes multiple images.

[0053] The first control module is used to perform a uniform speed operation to move away from the power transmission line according to the drone's positioning if the first target image does not identify the power tower, until the power tower is identified in the first target image.

[0054] The second control module is used to perform a reverse horizontal rotation based on the position of the power tower in the first target image if the first target image detects a power tower, until a set of power towers appears in the first target image; a set of power towers includes two power towers.

[0055] The composition calibration module is used to perform composition calibration based on the position of the group of power towers in the first target image, and then capture the video stream after calibration.

[0056] In one implementation, by acquiring real-time image data of the power transmission line (a front view of the power transmission line), the system can automatically identify and detect power towers in the first target image of the target image set using image recognition technology (at this point, the drone can identify the power transmission line but may not be able to detect the power tower). The drone takes off vertically towards the power transmission line. The automated execution of this step reduces the tediousness and time consumption of manual inspections, significantly improving inspection efficiency. When no power tower is identified in the first target image, it means that the distance to the power tower and the power transmission line is too close, and the sampling points for the power transmission line are too few. Therefore, the system controls the drone to move away from the power transmission line at a constant speed (reverse movement). That is, first determine the drone's own position, and then move away from the power transmission line according to the drone's position to prevent target loss, until the power tower is successfully identified in the image, ensuring the accuracy and comprehensiveness of the inspection.

[0057] In one implementation, after successfully identifying a power tower in the first target image, the system performs a reverse horizontal rotation based on the tower's position in the image. For example, if a power tower appears on the left side of the first target image, the system rotates in the opposite direction, meaning the drone's gimbal rotates to the right. This rotation only involves the horizontal direction. When two (or a group of) power towers appear in the image, identifying the group of towers allows for a more accurate determination of the transmission line's location and direction. The system then performs composition calibration based on the group of towers' positions in the image—that is, the drone's gimbal is fine-tuned—to center the transmission line in the image, with the group of towers positioned on either side. This improves the accuracy and stability of the image capture.

[0058] In one embodiment, the first control module includes: an image acquisition module, a feature fusion module, a coordinate determination module, a UAV positioning module, and an altitude determination module.

[0059] The image acquisition module is used to acquire measured images and satellite images captured by the UAV, and to randomly cut the measured images into sub-image sets according to a preset size;

[0060] The feature fusion module is used to extract feature maps from measured images, satellite images, and sub-image sets through the backbone network to obtain a feature map set, and then fuse the feature maps in the feature map set to obtain the final feature map;

[0061] The coordinate determination module is used to generate a target heatmap based on the final feature map and determine the UAV coordinates based on the heat distribution results in the target heatmap.

[0062] The UAV positioning module is used to interpolate the feature maps in the feature map set to obtain the target satellite image, and to determine the position of the UAV in the target satellite image based on the UAV coordinates;

[0063] The altitude determination module is used to determine the drone's altitude based on the measured image's position in the satellite image and the scaling ratio.

[0064] In one implementation, the measured image is a top-down view of the ground taken by a UAV. The backbone network receives the measured image taken by the UAV, satellite images, and multiple sub-images randomly cropped from the measured image. These are segmented into patches, and features are extracted using a Transformer encoder. Position encoding is used to enhance the feature representation. After feature extraction, a feature pyramid structure is used to process the feature map of the satellite image to retain semantic information at different scales. The feature map is then fused with the feature map of the UAV image to generate a heatmap. The feature map is restored to the size of the satellite image through interpolation to determine the relative position of the UAV on the satellite view.

[0065] In one implementation, a heatmap is generated by mapping the activation values ​​of feature maps to colors. It's typically used to visualize the model's attention or activation regions. Feature maps from a drone image, multiple sub-images, and a satellite image are fused. This fusion can be achieved in various ways, such as weighted summation or feature connection. The activation value of each cell (pixel) in the fused feature map is mapped to a color; the larger the activation value, the darker the color (e.g., red), indicating that the area is of focus for the model, resulting in the heatmap. The value of each pixel represents the probability or confidence that the location is a drone-captured image. "Hot" areas in the heatmap (areas with high probability or confidence) indicate that the model believes the drone-captured image is more likely to be located at that location, i.e., the drone image's position in the satellite image. The feature maps in the feature map set are interpolated to obtain the target satellite image. Interpolation methods include nearest neighbor interpolation, linear interpolation, bilinear interpolation, and bicubic interpolation. After determining the drone's height based on the scaling ratio, the drone's height can be adjusted to center the subsequent power transmission line within the image.

[0066] In one embodiment, the curve fitting module includes: a historical data acquisition module, a data training module, a model update module, and a feature extraction module.

[0067] The historical data acquisition module is used to acquire historical spacer image sets and historical transmission line image sets, and uses the historical spacer image sets and historical transmission line image sets as inputs for training a convolutional neural network model; the historical transmission line images include historical spacer images;

[0068] The data training module is used to extract historical spacer images and historical transmission line images through the feature extraction network to obtain feature maps, determine candidate regions in the feature maps through the region candidate network to obtain candidate region feature maps, and input the candidate region feature maps into the head network through the ROIAlign layer for evaluation to obtain the final candidate regions.

[0069] The model update module is used to use the final candidate region as the model update parameter, and update each layer in the convolutional neural network model according to the model update parameter to obtain the feature extraction model.

[0070] The feature extraction module is used to divide the video stream according to a preset period to obtain a third target image, and input the third target image into the feature extraction model for training to obtain line features; the line features contain multiple feature points.

[0071] In one implementation, by acquiring historical spacer bar image sets and historical transmission line image sets, and using these as inputs to train a convolutional neural network model, the system can learn the complex features of the transmission line and its spacers. Using spacers as feature points for identification can significantly improve the model's recognition accuracy. The spacers can also be used to determine the location of the transmission line. Spacers are hardware (widely used iron or aluminum metal accessories in transmission lines) installed on split conductors to fix the spacing between them, preventing conductor collisions and suppressing aerobatic vibrations and span oscillations. Spacers are typically installed in the middle of the span, spaced 50–60 meters apart.

[0072] In one implementation, during training, the system extracts feature maps from historical images using a feature extraction network, determines candidate regions using a region candidate network, and then inputs the candidate region feature maps into the head network for evaluation through a ROI Alliance layer. This process considers not only the overall features of the transmission line but also the local features of key components, thereby improving the accuracy of recognition. The convolutional neural network model has strong generalization capabilities, enabling the system to cope with complex environments and changes in lighting conditions while maintaining stable recognition performance.

[0073] In one implementation, the system divides the video stream into third target images according to a preset period, and then inputs these images into a trained feature extraction model for real-time processing. Because convolutional neural network models have high computational efficiency and fast inference speed, the system can complete the recognition and analysis of a large number of images in a short time, reducing working time and optimizing work efficiency.

[0074] In one embodiment, the curve fitting module is further configured to determine the center coordinates of each feature point based on the feature points in the line features, fit the center coordinates of each feature point to obtain the spatial curve equation through a spatial fitting algorithm, and calculate the first sag value.

[0075] In one implementation, a convolutional neural network model is trained on a historical image set to extract transmission line features (feature points). A two-dimensional curve equation is then fitted using the center coordinates of these feature points. The center coordinates refer to the specific location of the feature point in space (image), determined by its x and y values ​​in a two-dimensional coordinate system—essentially establishing a Cartesian coordinate system. The spatial fitting algorithm is a mathematical method that analyzes a set of data points (such as the center coordinates of feature points) to find an optimal mathematical model (such as a spatial curve equation) to describe the overall distribution or trend of these data points. Using this algorithm, the center coordinates of the feature points can be fitted into a spatial curve, thus obtaining the equation of that curve. In the above steps, based on the precise location information of the line feature points, a spatial curve equation describing the line morphology is automatically generated using mathematical methods (spatial fitting algorithms). This improves the accuracy of the transmission line morphology description and also allows for further calculation of the first sag value.

[0076] In one embodiment, calculating the second sag value based on attribute features includes: a multi-parameter curve equation fitting module and a second sag value calculation module.

[0077] The multi-parameter curve equation fitting module is used to acquire the attribute characteristics of transmission lines in real time, and to obtain multi-parameter curve equations by fitting curve equations based on the attribute characteristics.

[0078] The second sag numerical calculation module is used to calculate the second sag value based on the multi-parameter curve equation.

[0079]

[0080] Where f0 is the sag value, g is the specific load, L is the horizontal distance between the two suspension points, k0 is the tension of the conductor at the suspension point, α is the coefficient of thermal expansion of the conductor, and ΔT is the change in temperature.

[0081] In one implementation, the specific load is the weight per unit length of conductor. By acquiring the attribute characteristics of the transmission line in real time and using these characteristics to perform high-precision curve equation fitting, the system can automatically generate a multi-parameter curve equation containing multiple parameters, thereby achieving real-time and accurate calculation of the sag value. This significantly improves the intelligence level of transmission line monitoring, making monitoring work more convenient and efficient, and effectively enhancing the safety and stability of the power system.

[0082] In one implementation, the weight per unit length of the conductor can be obtained through laboratory measurement or by consulting data provided by the manufacturer. In actual operation, if the conductor does not undergo significant changes (such as corrosion, wear, or changes in additional load), its weight per unit length can be considered constant. The horizontal distance between the two suspension points can be obtained through actual measurement or by consulting design documents, which are typically determined in design drawings at the beginning of transmission line construction. The tension of the conductor at the suspension point can be measured in real time using a tension sensor or strain gauge, which can be installed on the conductor or attached to the suspension point. The changes in tension and temperature, as well as the coefficient of thermal expansion of the conductor, are monitored. Temperature can be measured in real time using temperature sensors (such as thermistors and thermocouples). The coefficient of thermal expansion is usually obtained through laboratory testing and is an inherent property of the conductor material. For example: Copper wire has a coefficient of thermal expansion of approximately 1.7 × 10⁻⁵ / ℃ (or 0.000017 / ℃). When the temperature of copper wire increases by 1℃, its length will increase by approximately 0.000017 times its original length. Steel wire has a coefficient of thermal expansion of approximately 1.2 × 10⁻⁵ / ℃ (or 0.000012 / ℃). Compared to copper wire, steel wire has a smaller coefficient of thermal expansion. Aluminum wire has a coefficient of thermal expansion of approximately 2.3 × 10⁻⁵ / ℃ (or 0.000023 / ℃). Aluminum wire has a relatively large coefficient of thermal expansion and is more sensitive to temperature changes.

[0083] In one embodiment, the transmission line fault early warning system includes: a risk level assessment module, a first early warning module, and a second early warning module.

[0084] The risk level assessment module is used to upload the fault location of the transmission line to the system and conduct a risk level assessment if the absolute difference between the first sag value and the second sag value is greater than the risk threshold.

[0085] The first early warning module is used to increase the monitoring frequency of the transmission line within a preset time period if the risk threshold is less than the absolute difference and less than the high risk threshold.

[0086] The second early warning module is used to issue a fault alarm and upload the monitoring data of the transmission line to the cloud if the high-risk threshold is less than the absolute difference.

[0087] In one implementation, timely detection and handling of transmission line faults in a power system is crucial for ensuring stable grid operation. Through the steps described above, when the absolute difference between the first and second sag values ​​exceeds a risk threshold, the system can quickly locate the fault and assess its risk level. This automated processing method can shorten fault location time, improve operation and maintenance efficiency, and help maintenance personnel take appropriate measures based on different risk levels, reducing equipment damage.

[0088] In one implementation, the condition of transmission lines may be affected by various factors, such as weather and load. Through the above steps, when the risk threshold is less than the absolute difference but less than the high-risk threshold, the system can increase the monitoring frequency of the transmission lines within a preset time period. This adjustment not only enables the system to perform targeted monitoring based on different situations but also improves the efficiency and accuracy of monitoring.

[0089] In one implementation, the safety and reliability of transmission lines are directly related to the stable operation of the entire power grid. Through the steps described above, when the high-risk threshold is less than the absolute difference, the system will issue a fault alarm and upload monitoring data to the cloud. This alarm and data processing method enables maintenance personnel to quickly understand the fault situation and take effective measures to handle it. Furthermore, through data sharing and analysis in the cloud, the safety and reliability of the entire power system are improved. For example, in the event of a serious fault, the system can immediately issue an alarm and upload detailed monitoring data to the cloud, allowing maintenance personnel and other relevant departments to respond quickly, collaborate in handling the fault, and ensure the stable operation of the power grid.

[0090] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. An online monitoring system for conductor sag in transmission lines, characterized in that, The system includes: a positioning calibration module, a curve fitting module, a sag calculation module, and a fault early warning module. The positioning calibration module is used to send a power line locking command to enable the UAV to position and calibrate, and to acquire a video stream of the power line in the work area; The curve fitting module is used to extract features from the video stream image to obtain line features, and to fit the line features to obtain a space curve equation. The sag calculation module is used to calculate the first sag value according to the spatial curve equation, acquire the attribute characteristics of the transmission line in real time, and calculate the second sag value according to the attribute characteristics. The fault warning module is used to issue a transmission line fault warning if the absolute difference between the first sag value and the second sag value is greater than a risk threshold.

2. The online monitoring system for conductor sag of a transmission line according to claim 1, characterized in that, The positioning calibration module includes: a power tower identification module, a first control module, and a second control module; the power tower identification module is used to acquire image data of the power transmission line in real time to obtain a first target image set, and to perform power tower identification and detection on the first target image in the target image set; the target image set includes multiple images. The first control module is used to perform a uniform speed operation to move away from the power transmission line according to the drone's positioning if the first target image does not identify the power tower, until the power tower is identified in the first target image; The second control module is used to, if the first target image identifies a power tower, perform a reverse horizontal rotation based on the position of the power tower in the first target image until a set of power towers appears in the first target image; a set of power towers includes two power towers.

3. The online monitoring system for conductor sag of a transmission line according to claim 2, characterized in that, The positioning calibration module also includes a composition calibration module, which is used to perform composition calibration based on the position of the group of power towers in the first target image, and then capture a video stream after calibration.

4. The online monitoring system for conductor sag of a transmission line according to claim 2, characterized in that, The first control module includes: an image acquisition module, a feature fusion module, a coordinate determination module, a UAV positioning module, and an altitude determination module; The image acquisition module is used to acquire measured images and satellite images captured by the UAV, and to randomly cut the measured images into sub-image sets according to a preset size; The feature fusion module is used to extract feature maps from measured images, satellite images, and sub-image sets through the backbone network to obtain a feature map set, and to fuse the feature maps in the feature map set to obtain the final feature map; The coordinate determination module is used to generate a target heat map based on the final feature map, and to determine the UAV coordinates based on the heat distribution results in the target heat map. The UAV positioning module is used to interpolate the feature maps in the feature map set to obtain a target satellite image, and to determine the position of the UAV in the target satellite image based on the UAV coordinates; The altitude determination module is used to determine the drone altitude based on the position of the measured image in the satellite image and the scaling ratio, and based on the scaling ratio.

5. The online monitoring system for conductor sag of a transmission line according to claim 1, characterized in that, The curve fitting module includes: a historical data acquisition module, a data training module, a model update module, and a feature extraction module; The historical data acquisition module is used to acquire a set of historical spacer images and a set of historical transmission line images, and to use the set of historical spacer images and the set of historical transmission line images as input to a convolutional neural network model for training; the historical transmission line images include historical spacer images; The data training module is used to extract historical spacer images and historical transmission line images through a feature extraction network to obtain feature maps, determine candidate regions in the feature maps through a region candidate network to obtain candidate region feature maps, and input the candidate region feature maps into the head network through the ROIAlign layer for evaluation to obtain the final candidate regions. The model update module is used to use the final candidate region as the model update parameter, and update each layer in the convolutional neural network model according to the model update parameter to obtain the feature extraction model. The feature extraction module is used to divide the video stream according to a preset period to obtain a second target image, and input the second target image into the feature extraction model for training to obtain line features; the line features include multiple feature points.

6. The online monitoring system for conductor sag of a transmission line according to claim 1, characterized in that, The curve fitting module is also used to determine the center coordinates of each feature point based on the feature points in the line features, fit the center coordinates of each feature point to obtain the spatial curve equation through a spatial fitting algorithm, and calculate the first sag value.

7. The online monitoring system for conductor sag of a transmission line according to claim 1, characterized in that, The sag calculation module includes: a multi-parameter curve equation fitting module and a second sag numerical calculation module. The multi-parameter curve equation fitting module is used to acquire the attribute characteristics of the transmission line in real time, and to obtain the multi-parameter curve equation by fitting the curve equation based on the attribute characteristics. The second sag value calculation module is used to calculate the second sag value based on the multi-parameter curve equation; Where f0 is the sag value, g is the specific load, L is the horizontal distance between the two suspension points, k0 is the tension of the conductor at the suspension point, α is the coefficient of thermal expansion of the conductor, and ΔT is the change in temperature.

8. The online monitoring system for conductor sag of a transmission line according to claim 1, characterized in that, The fault early warning module includes a risk level assessment module, which is used to upload the fault location of the transmission line to the system and perform a risk level assessment if the absolute difference between the first sag value and the second sag value is greater than the risk threshold.

9. The online monitoring system for conductor sag of a transmission line according to claim 8, characterized in that, The fault early warning module also includes a first early warning module, which is used to increase the monitoring frequency of the transmission line within a preset time period if the risk threshold is less than the absolute difference value and less than the high risk threshold.

10. The online monitoring system for conductor sag of a transmission line according to claim 8, characterized in that, The fault warning module also includes a second warning module, which is used to issue a fault alarm and upload the monitoring data of the transmission line to the cloud if the high-risk threshold is less than the absolute difference.

Images