Power transmission line galloping monitoring method and device based on unmanned aerial vehicle, computer equipment, storage medium and program product
By using drones to collect video images and utilizing a target detection network, the problem of fixed cameras being unable to provide comprehensive coverage and accurate monitoring of power transmission line galloping has been solved, enabling accurate galloping monitoring and real-time alarms in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-04-10
AI Technical Summary
In existing technologies, visual monitoring methods using fixed cameras are difficult to achieve comprehensive coverage monitoring of large-scale power transmission lines, and are also difficult to accurately monitor the galloping state of power transmission lines in complex environments.
The method involves using drones to collect target video images, extracting image features using a target detection network, and determining the galloping data of the line through target tracking. This includes image detection, target tracking, and galloping data determination.
It enables accurate monitoring of transmission line galloping in complex environments, and can track and display the line's motion status in real time, providing information such as galloping amplitude and frequency.
Smart Images
Figure CN121838009A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of power equipment monitoring, and in particular to a galloping monitoring method and device for a power transmission line based on a UAV, a computer device and a program product. BACKGROUND
[0002] Galloping of a power transmission line is prone to occur under severe weather conditions such as strong wind, and this low-frequency large-amplitude vibration can cause conductor fatigue damage, fitting damage, interphase flashover and even tower collapse accidents. For a long time, the power industry has mainly relied on contact sensors for galloping monitoring. These sensors can provide direct measurement data, but have inherent limitations such as difficult installation and maintenance, limited monitoring points, and high cost.
[0003] In related technologies, the galloping state of a power transmission line is generally monitored by a visual monitoring method of a fixed camera. However, this method is limited by the monitoring field of view and the installation position, and it is difficult to achieve comprehensive coverage monitoring of a large range of power transmission lines. Moreover, in a complex monitoring environment, it is also difficult to accurately monitor the galloping state of a power transmission line. SUMMARY
[0004] Therefore, the present application provides a galloping monitoring method and device for a power transmission line based on a UAV, a computer device and a program product to solve the problem that the galloping state of a power transmission line cannot be accurately monitored by a visual monitoring method of a fixed camera, and it is difficult to achieve comprehensive coverage monitoring of a large range of power transmission lines.
[0005] In a first aspect, the present application provides a galloping monitoring method for a power transmission line based on a UAV, which comprises: collecting a target video image of a target line at a target monitoring position by a UAV, wherein the target video image is composed of a video stream of a sequence of continuous multiple images; extracting target image features corresponding to the target video image by using a target detection network under multiple conditions of the target line, to obtain an image detection result of the target line according to the target video image; locating a target tracking result of the target line according to the image detection result of the target line; determining a displacement sequence data of the target line in the vertical direction with respect to time according to the target tracking result of the target line; determining target galloping data of the target line according to the displacement sequence data.
[0006] In some optional embodiments, the galloping monitoring method for a power transmission line based on a UAV in the present application further comprises: Based on the image detection results, target tracking results, and target dancing data, render the target video image and display the current motion state of the target line.
[0007] In some optional implementations, based on the target video image, a target detection network is used to extract target image features corresponding to the target video image in the case of multiple transmission lines of the target line, to obtain the image detection result of the target line, including: When the target line is a transmission line with spacers, the target detection network introduces a weighted bidirectional feature pyramid network into the neck network to perform multi-scale feature fusion on the target video image. In the detection head, a classification branch is used to track the class confidence of each position and a regression branch is used to track the bounding box coordinates of each position. A convolutional block attention mechanism module is added to extract the target image features of the target video image. The target image features characterize the feature information of the spacers. When the target line is a transmission line without spacers, the target detection network introduces residual blocks as the backbone network in the encoder to encode the target video image, adds attention gate control in the skip connection to suppress interference information in the target video image, and adds a spatial pyramid pooling module in the decoder to segment the target video image, thereby obtaining the target image features of the target video image. The target image features characterize the feature information of the transmission line.
[0008] In some optional implementations, when the target line is a transmission line without spacers, the target detection network optimizes itself using a target loss function during the detection of the target video image. The target loss function is calculated using the following formula:
[0009] in, Let be the target loss function. This is the first hyperparameter used to balance the weights of various losses. This is the second hyperparameter used to balance the weights of various losses. For the first loss function, For the second loss function, The total number of pixels in the target video image. The first detection by the object detection network The probability that a pixel belongs to a line. For the first The actual label value per pixel. It is a smoothing constant. As a weighting factor used to balance foreground and background, For focusing parameters.
[0010] In some optional embodiments, the target tracking result of the target line is located according to the image detection result of the target line, and the target tracking result of the target line comprises: According to the image detection result of the target line, the target position information of each frame image of the detection target of the target line is determined; A future trajectory list of the target line is created, and the future trajectory list records the identification information and the tracking position information of each frame image of the detection target of the target line; For each frame image of the detection target of the target line, the Euclidean distance between the target position information and the tracking position information of the detection target is calculated; For each frame image of the detection target of the target line, the frame image with the Euclidean distance greater than a preset threshold is taken as the target tracking result of the target line, and the identification information corresponding to the frame image is also marked in the target tracking result of the target line.
[0011] In some optional embodiments, the target dance data comprises a target dance amplitude, and the target dance data of the target line is determined according to the displacement sequence data, comprising: The maximum displacement data and the minimum displacement data are determined from the displacement sequence data; The difference between the maximum displacement data and the minimum displacement data is calculated; The difference is taken as the target dance amplitude of the target line.
[0012] In some optional embodiments, the power transmission line dance monitoring method based on the unmanned aerial vehicle in the embodiment further comprises: When the target dance amplitude exceeds a preset threshold, an alarm action is performed.
[0013] In some optional embodiments, the target dance data comprises a target dance frequency, and the target dance data of the target line is determined according to the displacement sequence data, comprising: The displacement sequence data is filtered and denoised; The displacement sequence data after the filtering and denoising is analyzed by using a zero-crossing point detection algorithm to obtain the target dance frequency of the target line. In a second aspect, the embodiment of the present application provides a power transmission line dance monitoring device based on an unmanned aerial vehicle, and the device comprises: An image data acquisition module is configured to acquire target video images of a target line at a target monitoring position by using an unmanned aerial vehicle, wherein the target video images are composed of a video stream of a plurality of continuous frame images; An image data detection module is configured to extract target image features corresponding to the target video images by using a target detection network under the condition of a plurality of power transmission lines of the target line according to the target video images, so as to obtain an image detection result of the target line; An image data positioning module is configured to position a target tracking result of the target line according to the image detection result of the target line; A displacement data determination module is configured to determine displacement sequence data of the target line in a vertical direction over time according to the target tracking result of the target line. A galloping data determination module is configured to determine target galloping data of the target line according to the displacement sequence data.
[0014] In a third aspect, an embodiment of the present application provides a computer device, comprising: A memory and a processor, which are communicatively connected, and the memory stores computer instructions, and the processor executes the computer instructions to perform the galloping monitoring method of the power transmission line based on the unmanned aerial vehicle in the first aspect or any of the implementation manners of the first aspect.
[0015] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores computer instructions, and the computer instructions are used to make a computer execute the galloping monitoring method of the power transmission line based on the unmanned aerial vehicle in the first aspect or any of the implementation manners of the first aspect.
[0016] In a fifth aspect, an embodiment of the present application provides a computer program product, which comprises computer instructions, and the computer instructions are used to make a computer execute the galloping monitoring method of the power transmission line based on the unmanned aerial vehicle in the first aspect or any of the implementation manners of the first aspect.
[0017] The technical scheme of the present application has the following advantages: The present application collects target video images of the target line at a target monitoring position by using an unmanned aerial vehicle, wherein the target video images are composed of a video stream of a plurality of continuous image sequences; an image detection result of the target line is obtained by using a target detection network to detect the target video images; a target tracking result of the target line is positioned according to the image detection result of the target line; displacement sequence data of the target line in a vertical direction over time is determined according to the target tracking result of the target line; and target galloping data of the target line is determined according to the displacement sequence data. Since the target detection network is used to detect the target video images, and the detection result is further positioned and tracked, the galloping information related to the target line is finally determined according to the tracking result. Therefore, the present application can accurately monitor the galloping state of the power transmission line even in a complex monitoring environment. BRIEF DESCRIPTION OF DRAWINGS
[0018] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.
[0019] Figure 1 is a flowchart of a method for monitoring the galloping of a power transmission line based on a UAV according to an embodiment of the present application; Figure 2 is a structural block diagram of a device for monitoring the galloping of a power transmission line based on a UAV according to an embodiment of the present application; Figure 3 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION
[0020] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.
[0021] According to an embodiment of the present application, a method for monitoring the galloping of a power transmission line based on a UAV is provided. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0022] In the present embodiment, a method for monitoring the galloping of a power transmission line based on a UAV is provided, which can be used in the above-mentioned computer device. The computer device can be a server, Figure 1 is a flowchart of a method for monitoring the galloping of a power transmission line based on a UAV according to an embodiment of the present application, as Figure 1 shown, the flow includes the following steps: Step S101, collecting a target video image of a target line at a target monitoring position by a UAV, wherein the target video image is a video stream composed of a sequence of continuous multiple images.
[0023] Specifically, the target monitoring position can be a tower section position of the power transmission line. For example, by patrolling the power transmission line by the unmanned aerial vehicle, the unmanned aerial vehicle is controlled to hover or directionally cruise at a key monitoring tower section of the power transmission line, and a target video image is obtained by performing spot video acquisition on the target monitoring position, the target video image being composed of a video stream of a continuous sequence of multiple frames.
[0024] In some optional embodiments, the method for monitoring galloping of the power transmission line based on the unmanned aerial vehicle in the embodiment includes: Step a1, obtaining a target attitude angle of the unmanned aerial vehicle.
[0025] Step a2, performing image enhancement processing on the target video image according to the target attitude angle.
[0026] For example, the target attitude angle of the unmanned aerial vehicle is collected by using an inertial sensor, the target attitude angle including a pitch angle, a roll angle and a yaw angle, and a rotation matrix is formed by the target attitude angles. It is assumed that the image after the enhancement processing is under ideal orthographic projection, and is then projected onto a new image plane by a camera intrinsic parameter. According to the rotation matrix and the camera intrinsic parameter, a homography matrix from the original oblique image to the image after the enhancement processing is determined, and finally the original image is inversely mapped according to the homography matrix, and an image after the enhancement processing is generated by an interpolation algorithm.
[0027] Step S102, according to the target video image, using a target detection network to extract target image features corresponding to the target video image in the case of multiple power transmission lines of the target line, and obtaining an image detection result of the target line.
[0028] In some specific embodiments, the step S102 of extracting the target image features corresponding to the target video image in the case of multiple power transmission lines of the target line according to the target video image by using the target detection network to obtain the image detection result of the target line includes: In the case of the target line being a power transmission line with a spacer, the target detection network introduces a weighted bidirectional feature pyramid network in a neck network to perform multi-scale feature fusion on the target video image, uses a classification branch in a detection head to track a class confidence of each position and uses a regression branch to track a bounding box coordinate of each position, and adds a convolution block attention mechanism module to extract target image features of the target video image, the target image features representing feature information of the spacer; In the case that the target line is a power transmission line without a spacer, the target detection network introduces a residual block as a backbone network in the encoder to encode the target video image, adds an attention gate control in the jump connection to suppress the interference information of the target video image, and adds a spatial pyramid pooling module in the decoder to segment the target video image, to obtain a target image feature of the target video image, which represents the feature information of the power transmission line.
[0029] For example, when the target line is a power transmission line with a spacer, and the target detection network is a YOLOv11 deep learning network, the target detection network automatically detects the spacer of the power transmission line in the input image. By introducing a multi-scale feature fusion mechanism and an attention module, the recognition accuracy of the spacer in a complex background is significantly improved. The specific improvements of the target detection network in this embodiment include: introducing a weighted bidirectional feature pyramid network (BiFPN network) in the neck network to achieve efficient multi-scale feature fusion; using a decoupled head design in the detection head part to optimize the classification and regression tasks respectively, i.e. using a classification branch to track the class confidence of each position and a regression branch to track the bounding box coordinates of each position. By adding a convolutional block attention module (CBAM), a first target feature of the target video image is extracted, which represents the feature information of the spacer. In terms of data augmentation, in addition to the conventional geometric transformation and color adjustment, motion blur and perspective change specific to unmanned aerial vehicle shooting are also simulated, so that the target detection network has better adaptability to the actual inspection environment.
[0030] In the training process of the target detection network, Mosaic data augmentation and adaptive anchor box calculation are used to further improve the detection performance of the model in complex scenes. Mosaic data augmentation is a data augmentation technique that combines four training images into one composite image in the above-mentioned YOLOv11 deep learning network. The image detection result of the target line output by the target detection network includes the horizontal coordinate, vertical coordinate, length, width, center point coordinate and confidence score of the target detection box.
[0031] For transmission lines without spacers, the target detection network employs a semantic segmentation network based on the U-Net deep learning model to automatically detect and extract transmission conductors from the input image. This embodiment improves upon U-Net by introducing a residual block (ResNet34) pre-trained on the ImageNet image dataset as the backbone in the encoder to enhance feature extraction capabilities; adding an attention gating mechanism to skip connections to effectively suppress interference from irrelevant background information; and incorporating a Spatial Pyramid Pooling (ASPP) module in the decoder to improve the segmentation performance of the target detection network for conductors at different scales.
[0032] In some specific implementations, when the target line is a transmission line without spacers, the target detection network optimizes itself using a target loss function during the detection of the target video image. The target loss function is calculated using the following formula:
[0033] in, Let be the target loss function. This is the first hyperparameter used to balance the weights of various losses. This is the second hyperparameter used to balance the weights of various losses. For the first loss function, For the second loss function, The total number of pixels in the target video image. The first detection by the object detection network The probability that a pixel belongs to a line. For the first The actual label value per pixel. It is a smoothing constant. As a weighting factor used to balance foreground and background, For focusing parameters.
[0034] In the above formula, and It is a hyperparameter used to balance the weights of various losses, and its value is... , , It is a very small smoothing constant to prevent the denominator from being zero. This represents the weighting factor used to balance the foreground and background, and can be set as the percentage of the wire pixels in the entire image. To focus on parameters, the weight of easily classified samples is reduced. Based on this, the extracted target transmission line is divided into 5 segments, with the starting position of the contour of each segment used as a key point. This segmentation strategy allows for in-depth analysis of the characteristics of the target line, ensuring a sufficient number of key points while avoiding computational redundancy caused by over-segmentation.
[0035] Step S103, according to the image detection result of the target line, the target tracking result of the target line is located.
[0036] In some specific embodiments, according to the image detection result of the target line, the target tracking result of the target line is located, including: Step b1, according to the image detection result of the target line, the target position information of each frame image of the detection target of the target line is determined.
[0037] For example, the target position information of the first frame image of the detection target of the target line is . .
[0038] Step b2, creating a future trajectory list of the target line, the future trajectory list records the identification information and tracking position information of each frame image of the detection target of the target line.
[0039] For example, the future trajectory list of the target line is: .
[0040] Step b3, for each frame image of the detection target of the target line, the Euclidean distance between the target position information and the tracking position information of the detection target is calculated.
[0041] For example,
[0042] Step b4, for each frame image of the detection target of the target line, the frame image with the Euclidean distance greater than the preset threshold is taken as the target tracking result of the target line, and the identification information corresponding to the frame image is also marked in the target tracking result of the target line.
[0043] According to the above example, the Euclidean distance between the target position information and the tracking position information of each frame image of the detection target of the target line is calculated, and the frame image with the Euclidean distance greater than the preset threshold is counted as the target tracking result of the target line.
[0044] The above embodiment can realize real-time and accurate tracking of the same detection target of the target line.
[0045] Step S104, according to the target tracking result of the target line, the displacement sequence data of the target line changing with time in the vertical direction is determined.
[0046] Specifically, the target tracking result of the target line is the position information of the target line in different directions, and the displacement sequence data of the target line changing with time in the vertical direction is extracted.
[0047] Step S105, determining target galloping data of the target line according to the displacement sequence data.
[0048] In some specific embodiments, the target galloping data includes a target galloping amplitude, and the step S104 includes determining displacement sequence data of the target line in the vertical direction changing over time according to the target tracking result of the target line. Step c1, determining maximum displacement data and minimum displacement data from the displacement sequence data.
[0049] Step c2, calculating a difference between the maximum displacement data and the minimum displacement data.
[0050] Step c3, taking the difference as the target galloping amplitude of the target line.
[0051] The embodiment determines the target galloping amplitude of the target line through the movement information of the target line in the vertical direction.
[0052] In some optional embodiments, the galloping monitoring method of the power transmission line based on the unmanned aerial vehicle in the embodiment further includes: When the target galloping amplitude exceeds the preset threshold, an alarm action is performed.
[0053] For example, when the target galloping amplitude exceeds the preset threshold, the “GALLOPING!” alarm mechanism is automatically triggered to realize real-time early warning of the abnormally large galloping.
[0054] In some specific embodiments, the target galloping data includes a target galloping frequency, and the step of determining the target galloping data of the target line according to the displacement sequence data includes: Step d1, filtering and denoising the displacement sequence data.
[0055] Step d2, analyzing the displacement sequence data after the filtering and denoising by using a zero-crossing point detection algorithm to obtain the target galloping frequency of the target line. For example, the displacement sequence data is first filtered and denoised to optimize the data. After detecting and analyzing the preprocessed displacement sequence data by using the zero-crossing point detection algorithm, 8 continuous zero-crossing points are identified within a preset time period (3s), which means that there are 7 complete periods, and the total number of periods is 7. The target galloping frequency of the target line is represented by f, and f=7 / 3≈2.33 Hz, which means that the target galloping frequency of the target line is approximately 2.33 periods per second.
[0056] In some optional embodiments, the galloping monitoring method of the power transmission line based on the unmanned aerial vehicle in the embodiment further includes: According to the image detection result, the target tracking result and the target dancing data, a target video image is rendered and a current motion state of the target line is displayed.
[0057] For example, the image detection result, the target tracking result and the target dancing data of the target line are superimposed and rendered in real time in the original target video image. The target tracking result includes the identification information of the detected target and the attribute information of the target detection box. The target tracking result also represents the historical trajectory of the target line. This embodiment can intuitively display the motion state and the recognition result of the target line. In specific implementation, the target detection box is marked with a highlighted rectangular box, the historical trajectory is depicted by a continuous curve, and a dynamic label is updated and displayed on the side of the picture in real time, including the identification information of the detected target, the target dancing amplitude and the target dancing frequency, so that the operator can clearly and efficiently monitor the dynamic change of the conductor and the detection effect of the system.
[0058] In this embodiment, a target video image of the target line is collected by a UAV at a target monitoring position, the target video image is a video stream composed of a sequence of continuous multiple images, the image detection result of the target line is obtained by using a target detection network to detect the target video image, the target tracking result of the target line is located according to the image detection result of the target line, the displacement sequence data of the target line in the vertical direction with time is determined according to the target tracking result of the target line, and the target dancing data of the target line is determined according to the displacement sequence data. Since the target detection network is used to detect the target video image in this embodiment, and the detection result is further located and tracked, the dancing information related to the target line is finally determined according to the tracking result. Therefore, the dancing state of the power transmission line can be accurately monitored even in a complex monitoring environment.
[0059] In this embodiment, a target video image of the target line is collected by a UAV at a target monitoring position, the target video image is a video stream composed of a sequence of continuous multiple images, the image detection result of the target line is obtained by using a target detection network to detect the target video image, the target tracking result of the target line is located according to the image detection result of the target line, the displacement sequence data of the target line in the vertical direction with time is determined according to the target tracking result of the target line, and the target dancing data of the target line is determined according to the displacement sequence data. Since the target detection network is used to detect the target video image in this embodiment, and the detection result is further located and tracked, the dancing information related to the target line is finally determined according to the tracking result. Therefore, the dancing state of the power transmission line can be accurately monitored even in a complex monitoring environment.
[0060] This embodiment provides a UAV-based power transmission line dancing monitoring device, as shown in Figure 2 , which includes: An image data acquisition module 201 is configured to collect a target video image of a target line at a target monitoring position by a UAV, wherein the target video image is a video stream composed of a sequence of continuous multiple images; The image data detection module 202 is configured to extract target image features corresponding to the target video image from the target video image by using a target detection network in the case of multiple power transmission lines of the target line, to obtain an image detection result of the target line. The image data positioning module 203 is configured to position a target tracking result of the target line according to the image detection result of the target line. The displacement data determination module 204 is configured to determine a displacement sequence data of the target line in the vertical direction over time according to the target tracking result of the target line. The galloping data determination module 205 is configured to determine target galloping data of the target line according to the displacement sequence data.
[0061] In some optional embodiments, the unmanned aerial vehicle-based galloping monitoring device for the power transmission line in the embodiment further includes: The monitoring data display module is configured to render the target video image and display a current motion state of the target line according to the image detection result, the target tracking result and the target galloping data.
[0062] In some optional embodiments, the image data detection module 202 is specifically configured to: In the case of the target line being a power transmission line with a spacer, the target detection network introduces a weighted bidirectional feature pyramid network in the neck network to perform multi-scale feature fusion on the target video image, uses a classification branch in the detection head to track the class confidence of each position and uses a regression branch to track the bounding box coordinates of each position, and adds a convolution block attention mechanism module to extract target image features of the target video image, the target image features representing feature information of the spacer; In the case of the target line being a power transmission line without a spacer, the target detection network introduces a residual block as a backbone network in the encoder to encode the target video image, adds an attention gate control in the skip connection to suppress interference information of the target video image, and adds a spatial pyramid pooling module in the decoder to segment the target video image, to obtain target image features of the target video image, the target image features representing feature information of the power transmission line.
[0063] In some optional embodiments, when the target line is a power transmission line without a spacer, the target detection network optimizes the target detection network by using a target loss function in the process of detecting the target video image, the target loss function being calculated by the following formula:
[0064] wherein, the target loss function is L, the first hyperparameter for balancing the loss weights is λ, a second hyper-parameter for balancing the loss weight of each item, a first loss function, a second loss function, a total number of pixels of the target video image, a probability that the first pixel detected by the target detection network belongs to the line, a true label value of the first pixel, a smoothing constant, a weight factor for balancing the foreground and the background, a focus parameter.
[0065] In some optional embodiments, the image data positioning module 203 is specifically configured to: determine target position information of each frame image of the detection target of the target line according to the image detection result of the target line; create a future trajectory list of the target line, the future trajectory list recording identification information and tracking position information of each frame image of the detection target of the target line; calculate the Euclidean distance between the target position information and the tracking position information of the detection target for each frame image of the detection target of the target line; for each frame image of the detection target of the target line, frame images with a Euclidean distance greater than a preset threshold value are taken as a target tracking result of the target line, and the identification information corresponding to the frame images is also marked in the target tracking result of the target line.
[0066] In some optional embodiments, the target dance data includes: a target dance amplitude, and the dance data determination module 205 is specifically configured to: determine maximum displacement data and minimum displacement data from the displacement sequence data; calculate a difference value between the maximum displacement data and the minimum displacement data; take the difference value as the target dance amplitude of the target line.
[0067] In some optional embodiments, the unmanned aerial vehicle-based transmission line dance monitoring device in the embodiment further includes: a data anomaly alarm module configured to perform an alarm action when the target dance amplitude exceeds a preset threshold value.
[0068] In some optional embodiments, the target dance data includes: a target dance frequency, and the dance data determination module 205 is specifically configured to: perform filtering and denoising processing on the displacement sequence data; analyze the displacement sequence data after the filtering and denoising processing by using a zero-crossing detection algorithm to obtain the target dance frequency of the target line. Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.
[0069] In this embodiment, the UAV-based transmission line galloping monitoring device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.
[0070] This invention also provides a computer device having the above-described features. Figure 2 The image shows a drone-based power line galloping monitoring device.
[0071] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 3 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 3 Take a processor 10 as an example.
[0072] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.
[0073] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.
[0074] The memory 20 can include a program storage area and a data storage area. The program storage area can store an operating system, application programs required for at least one function, etc. The data storage area can store data created by the computer device, etc. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 can optionally include memory that is remotely located with respect to the processor 10, and which can be connected to the computer device through a network. Examples of such networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communications network, and combinations thereof.
[0075] The memory 20 can include a volatile memory, such as a random access memory, and / or can include a non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage device. The memory 20 can also include an array of multi-state flash memory cells, which can be used to store data and / or instructions in multiple states.
[0076] The computer device also includes a communications interface 30 for communicating with other devices or communication networks.
[0077] The embodiments of the present application also provide a computer readable storage medium, and the method according to the embodiments of the present application can be implemented in hardware, firmware, or recorded in a storage medium, or implemented as computer code to be originally stored in a remote storage medium or a non-transitory machine readable storage medium downloaded through a network and stored in a local storage medium, so that the method described herein can be processed by such software on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state disk, etc. Further, the storage medium can also include a combination of the above-mentioned kinds of storage. It can be understood that the computer, processor, microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the method shown in the above embodiments.
[0078] Part of the present application can be applied as a computer program product, for example, computer program instructions, when executed by a computer, through the operation of the computer, can invoke or provide the method and / or technical solutions according to the present application. Those skilled in the art should understand that the form of computer program instructions in computer readable medium includes but is not limited to source files, executable files, installation package files and the like, and accordingly, the way of computer program instructions executed by computer includes but is not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Here, the computer readable medium can be any available computer readable storage medium or communication medium accessible to the computer.
[0079] Although the embodiments of the present application are described in conjunction with the drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.
Claims
1. A method for monitoring the galloping of power transmission lines based on unmanned aerial vehicles (UAVs), characterized in that, The method includes: The target video images of the target route are collected by a drone at the target monitoring location, wherein the target video images are composed of a video stream consisting of a continuous sequence of multiple frames; Based on the target video image, a target detection network is used to extract the target image features corresponding to the target video image under the condition of multiple transmission lines of the target line, so as to obtain the image detection result of the target line; Based on the image detection results of the target route, the target tracking results of the target route are located; Based on the target tracking results of the target line, determine the displacement sequence data of the target line in the vertical direction over time; Based on the displacement sequence data, the target galloping data of the target line is determined.
2. The method according to claim 1, characterized in that, The method further includes: Based on the image detection results, the target tracking results, and the target motion data, the target video image is rendered and the current motion state of the target line is displayed.
3. The method according to claim 1, characterized in that, Based on the target video image, a target detection network is used to extract target image features corresponding to the target video image under various transmission line conditions of the target line, to obtain the image detection result of the target line, including: In the case where the target line is a transmission line with spacers, the target detection network introduces a weighted bidirectional feature pyramid network into the neck network to perform multi-scale feature fusion on the target video image. In the detection head, a classification branch is used to track the class confidence of each position and a regression branch is used to track the bounding box coordinates of each position. A convolutional block attention mechanism module is added to extract the target image features of the target video image, which characterize the feature information of the spacers. When the target line is a transmission line without spacers, the target detection network introduces residual blocks as a backbone network in the encoder to encode the target video image, adds attention gate control in the skip connection to suppress interference information in the target video image, and adds a spatial pyramid pooling module in the decoder to segment the target video image to obtain the target image features of the target video image. The target image features characterize the feature information of the transmission line.
4. The method according to claim 1, characterized in that, When the target line is a transmission line without spacers, the target detection network optimizes itself using a target loss function during the detection of the target video image. The target loss function is calculated using the following formula: in, Let the target loss function be... This is the first hyperparameter used to balance the weights of various losses. This is the second hyperparameter used to balance the weights of various losses. For the first loss function, For the second loss function, The total number of pixels in the target video image. The first target detected by the target detection network The probability that a pixel belongs to a line. For the first The actual label value per pixel. It is a smoothing constant. As a weighting factor used to balance foreground and background, For focusing parameters.
5. The method according to claim 1, characterized in that, Based on the image detection results of the target route, the target tracking results for locating the target route are obtained, including: Based on the image detection results of the target line, determine the target position information of each frame of the detected target line; Create a future trajectory list for the target route, wherein the future trajectory list records the identification information and tracking position information of each frame of the detected target on the target route; For each frame of the image of the detected target on the target line, calculate the Euclidean distance between the target position information and the tracking position information of the detected target; For each frame of the target image of the target line, the frame image with a Euclidean distance greater than a preset threshold is taken as the target tracking result of the target line. The target tracking result of the target line is also marked with the identification information corresponding to the frame image.
6. The method according to claim 1, characterized in that, The target galloping data includes: target galloping amplitude. Based on the displacement sequence data, the target galloping data of the target line is determined, including: From the displacement sequence data, determine the maximum displacement data and the minimum displacement data; Calculate the difference between the maximum displacement data and the minimum displacement data; The difference is used as the target dancing amplitude of the target line.
7. The method according to claim 6, characterized in that, The method further includes: When the target's dancing amplitude exceeds a preset threshold, an alarm action is triggered.
8. The method according to claim 1, characterized in that, The target galloping data includes: target galloping frequency. Based on the displacement sequence data, the target galloping data of the target line is determined, including: The displacement sequence data is then filtered and denoised. By using a zero-crossing detection algorithm, the displacement sequence data after filtering and denoising is analyzed to obtain the target galloping frequency of the target line.
9. A device for monitoring the galloping of power transmission lines based on unmanned aerial vehicles (UAVs), characterized in that, The device includes: The image data acquisition module is used to acquire target video images of the target route at the target monitoring location using a drone, wherein the target video images are composed of a video stream consisting of a continuous sequence of multiple frames; The image data detection module is used to extract target image features corresponding to the target video image based on the target video image and to obtain the image detection result of the target line under various transmission line conditions of the target line using a target detection network. The image data localization module is used to locate the target tracking result of the target line based on the image detection result of the target line; The displacement data determination module is used to determine the displacement sequence data of the target line in the vertical direction over time based on the target tracking results of the target line; The dancing data determination module is used to determine the target dancing data of the target line based on the displacement sequence data.
10. A computer device, characterized in that, include: The system includes a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the method for monitoring the galloping of power transmission lines based on any one of claims 1 to 8.
11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the unmanned aerial vehicle-based transmission line galloping monitoring method according to any one of claims 1 to 8.
12. A computer program product, characterized in that, Includes computer instructions for causing a computer to execute the unmanned aerial vehicle-based transmission line galloping monitoring method according to any one of claims 1 to 8.