Breaker opening and closing speed determination method and device and storage medium

Abstract

The invention discloses a circuit breaker opening and closing speed determination method and device and a storage medium. The method comprises the steps that an initial action video of circuit breaker opening and closing is acquired, the circuit breaker opening and closing is used for representing the on-off state of a circuit where the circuit breaker is located, the initial action video is used for representing the process that an operating mechanism in the circuit breaker controls the circuit breaker to execute switching action, and the operating mechanism is an assembly in the circuit breaker; performing image enhancement processing on a plurality of initial images in the initial action video to obtain a plurality of target images, and generating a target action video of opening and closing of the circuit breaker from the plurality of target images, the plurality of target images being in one-to-one correspondence with the plurality of initial images; performing target tracking processing on the target action video to generate a moving track of the circuit breaker in the target action video; and based on the moving trajectory, determining a target speed of opening and closing of the circuit breaker. The technical problem that it is difficult to obtain the appropriate opening and closing speed of the circuit breaker is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit breakers, and in particular to a method, a device and a storage medium for determining the speed of opening and closing of a circuit breaker. Background Art

[0002] At present, the opening and closing speed of circuit breakers is an important indicator for evaluating the reliability of circuit breakers. The commonly used method to obtain the opening and closing speed of circuit breakers is to use special testing instruments, such as opening and closing speed testers or oscilloscopes, to measure the opening and closing time of circuit breakers to determine whether their speed meets the standard requirements.

[0003] However, if the circuit breaker opening and closing speed is too fast, it may cause excessive impact on the operating mechanism, thereby damaging the components; if it is too slow, it may cause the arcing time between the contacts to be prolonged, increasing electrical wear. As a result, it is difficult to obtain a suitable circuit breaker opening and closing speed.

[0004] With respect to the above-mentioned technical problem of difficulty in obtaining a suitable circuit breaker opening and closing speed, no effective solution has been proposed so far. Summary of the invention

[0005] The embodiments of the present invention provide a method, device and storage medium for determining the opening and closing speed of a circuit breaker, so as to at least solve the technical problem that it is difficult to obtain a suitable opening and closing speed of a circuit breaker.

[0006] According to one aspect of an embodiment of the invention, a method for determining the speed of circuit breaker opening and closing is provided. The method may include: obtaining an initial action video of circuit breaker opening and closing, wherein the circuit breaker opening and closing is used to characterize the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of the operating mechanism in the circuit breaker controlling the circuit breaker to perform the switching action, and the operating mechanism is a component in the circuit breaker; performing image enhancement processing on multiple initial images in the initial action video respectively to obtain multiple target images, and generating a target action video of circuit breaker opening and closing from the multiple target images, wherein the multiple target images correspond to the multiple initial images one by one; performing target tracking processing on the target action video to generate the running trajectory of the circuit breaker in the target action video; and determining the target speed of circuit breaker opening and closing based on the running trajectory.

[0007] According to one aspect of an embodiment of the present invention, a speed determination device for opening and closing of a circuit breaker is provided, the device comprising: a first acquisition unit, used to acquire an initial action video of opening and closing of the circuit breaker, wherein the opening and closing of the circuit breaker is used to characterize the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of an operating mechanism in the circuit breaker controlling the circuit breaker to perform a switching action, and the operating mechanism is a component in the circuit breaker; a second acquisition unit, used to perform image enhancement processing on multiple initial images in the initial action video respectively to obtain multiple target images, and generate a target action video of opening and closing of the circuit breaker from the multiple target images, wherein the multiple target images correspond one-to-one to the multiple initial images; a generation unit, used to perform target tracking processing on the target action video to generate a running trajectory of the circuit breaker in the target action video; and a determination unit, used to determine the target speed of opening and closing of the circuit breaker based on the running trajectory.

[0008] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the device where the storage medium is located is controlled to execute the method in the embodiment of the present invention.

[0009] According to another aspect of an embodiment of the present invention, a processor is provided. The processor is used to run a program, wherein the method in the embodiment of the present invention is executed when the program is run.

[0010] According to another aspect of the embodiments of the present invention, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the method in the embodiments of the present invention is implemented.

[0011] According to another aspect of an embodiment of the present invention, an electronic device is provided, the electronic device comprising a processor and a memory for storing instructions executable by the processor, wherein the processor is configured to execute the instructions to implement the method in the embodiment of the present invention.

[0012] In an embodiment of the present invention, an initial action video of the opening and closing of a circuit breaker is obtained, wherein the opening and closing of the circuit breaker is used to characterize the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of an operating mechanism in the circuit breaker controlling the circuit breaker to perform a switching action, and the operating mechanism is a component in the circuit breaker; image enhancement processing is performed on multiple initial images in the initial action video to obtain multiple target images, and the multiple target images are used to generate a target action video of the opening and closing of the circuit breaker, wherein the multiple target images correspond one-to-one to the multiple initial images; target tracking processing is performed on the target action video to generate a running trajectory of the circuit breaker in the target action video; and based on the running trajectory, a target speed of the opening and closing of the circuit breaker is determined. That is to say, the embodiment of the present invention can first obtain the initial action video of the circuit breaker opening and closing, and then perform image enhancement processing on multiple initial images in the initial action video respectively to obtain multiple target images, and generate the target action video of the circuit breaker opening and closing from the multiple target images, and then perform target tracking processing on the target action video to generate the operation trajectory of the circuit breaker in the target action video. Finally, according to the operation trajectory obtained above, the purpose of determining the target speed of the circuit breaker opening and closing is achieved, thereby solving the technical problem of difficulty in obtaining a suitable circuit breaker opening and closing speed, and achieving the technical effect of being able to obtain a suitable circuit breaker opening and closing speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0014] Figure 1 is a flow chart of a method for determining the speed of opening and closing of a circuit breaker according to an embodiment of the present invention;

[0015] Figure 2 is a flow chart of a circuit breaker opening and closing speed detection system based on drone image acquisition according to an embodiment of the present invention;

[0016] Figure 3 is a schematic diagram of a drone system according to an embodiment of the present invention;

[0017] Figure 4 is a schematic diagram of an image processing module according to an embodiment of the present invention;

[0018] Figure 5 is a schematic diagram of obtaining image data of circuit breaker opening and closing collected by a drone according to an embodiment of the present invention;

[0019] Figure 6 The invention relates to a circuit breaker opening and closing speed determining device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] According to an embodiment of the present invention, a method for determining the speed of opening and closing of a circuit breaker is provided. It should be noted that in the flowchart of the accompanying drawings, the steps shown therein can be executed in a computer system such as a set of computer executable instructions, and although a 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.

[0023] The following is an introduction to a method for determining the speed of opening and closing of a circuit breaker according to an embodiment of the present invention.

[0024] Figure 1 is a flow chart of a method for determining the speed of opening and closing of a circuit breaker according to an embodiment of the present invention. Figure 1 As shown, the method for securely storing the power grid privileged access credentials may include the following steps:

[0025] Step S101, obtaining an initial action video of the circuit breaker opening and closing.

[0026] In the technical solution provided in the above step S101 of the present invention, an initial action video of the circuit breaker opening and closing can be obtained, wherein the circuit breaker opening and closing is used to characterize the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of the operating mechanism in the circuit breaker controlling the circuit breaker to perform the switching action. The operating mechanism is a component in the circuit breaker, and the circuit breaker can be called a high-voltage circuit breaker.

[0027] Optionally, the opening and closing of the circuit breaker may include two operations of the circuit breaker: a closing operation of the circuit breaker and an opening operation of the circuit breaker. For example, after the closing operation is successfully performed on the circuit breaker, the circuit in which the circuit breaker is located is closed, and current can flow at this time; after the opening operation is successfully performed on the circuit breaker, the circuit in which the circuit breaker is located is disconnected, and current cannot flow at this time.

[0028] For example, after performing closing and opening operations on the circuit breaker, an initial action video of the above process is collected, and then the initial action video is further processed.

[0029] Step S102, performing image enhancement processing on multiple initial images in the initial action video respectively to obtain multiple target images, and generating a target action video of the circuit breaker opening and closing from the multiple target images.

[0030] In the technical solution provided in the above step S102 of the present invention, after obtaining the initial action video, image enhancement processing is performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the above-obtained multiple target images are used to generate a target action video of the circuit breaker opening and closing.

[0031] Optionally, the multiple target images correspond to the multiple initial images one by one. Image enhancement processing is used to improve the quality and visual effect of an image so as to better represent the information in the image and make the image more suitable for people's viewing and analysis needs.

[0032] For example, after capturing the initial action video of the circuit breaker opening and closing, image extraction is performed from the initial action video to obtain multiple initial images, and then image enhancement processing can be performed on the multiple initial images to obtain multiple target images. The multiple target images can form a target action video of the circuit breaker opening and closing.

[0033] It should be noted that this is only a preferred implementation method for generating a target action video of circuit breaker opening and closing, and the process and method for generating the target action video of circuit breaker opening and closing are not specifically limited. As long as image enhancement processing is performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the multiple target images are used to generate the target action video of circuit breaker opening and closing, the process and method are within the protection scope of the present invention and will not be repeated here.

[0034] Step S103: performing target tracking processing on the target action video to generate a running track of the circuit breaker in the target action video.

[0035] In the technical solution provided in the above step S103 of the present invention, based on the target action video obtained in the above step, target tracking processing can be performed on the target action video to achieve the purpose of generating the running trajectory of the circuit breaker in the target action video.

[0036] Optionally, the target action video is tracked using corner point detection and an improved Kalman filter (Kanade-Lucas-Tomasi, KLT for short) optical flow method, and a tracking curve is depicted based on the tracked corner points to obtain a travel curve of the circuit breaker. Based on the travel curve, the running trajectory of the circuit breaker in the target action video can be obtained.

[0037] Optionally, corner detection is used to characterize corners in an image, where corners are usually important feature points in an image and are also the points where the local pixel value changes to the maximum. The KLT optical flow method is used to estimate the direction and speed of movement of each pixel by analyzing the change in pixel grayscale values ​​between adjacent frames, and is often used in tasks such as target tracking and motion analysis.

[0038] It should be noted that this is only a preferred implementation method for generating the running trajectory of the circuit breaker in the target action video, and the process and method of generating the running trajectory of the circuit breaker in the target action video are not specifically limited. As long as the target action video is subjected to target tracking processing, the process and method of generating the running trajectory of the circuit breaker in the target action video are within the protection scope of the present invention and are not listed here.

[0039] Step S104, determining the target speed of opening and closing of the circuit breaker based on the operation trajectory.

[0040] In the technical solution provided in the above step S104 of the present invention, the target speed of opening and closing of the circuit breaker can be determined according to the obtained running trajectory of the circuit breaker in the target action video.

[0041] Optionally, the target speed can be the circuit breaker opening and closing speed, or the opening and closing speed, or the maximum speed, etc. It should be noted that the target speed is only illustrated here, and the speed meaning of the target speed is not specifically limited. As long as the target speed of circuit breaker opening and closing is obtained according to the operating trajectory, it is within the protection scope of the present invention.

[0042] For example, after obtaining the travel curve of the high-voltage circuit breaker by tracking the curve according to the tracking corner points, the position of the just-opening and closing points can be queried, and then the opening and closing speed, just-opening and closing speed, maximum speed, etc. of the high-voltage circuit breaker can be obtained.

[0043] In the above steps S101 to S104 of the present invention, an initial action video of the opening and closing of the circuit breaker can be first obtained, and then image enhancement processing can be performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the multiple target images are used to generate a target action video of the opening and closing of the circuit breaker, and then the target action video is subjected to target tracking processing to generate a running trajectory of the circuit breaker in the target action video. Finally, according to the running trajectory obtained above, the target speed of the opening and closing of the circuit breaker is determined, thereby solving the technical problem of difficulty in obtaining a suitable opening and closing speed of the circuit breaker and achieving the technical effect of obtaining a suitable opening and closing speed of the circuit breaker.

[0044] The above method of this embodiment is further introduced below.

[0045] As an optional implementation method, obtaining the initial action video of the circuit breaker opening and closing includes: receiving the initial action video of the circuit breaker opening and closing taken from a target drone, wherein the anti-electromagnetic interference capability of the receiver in the target drone is higher than the anti-electromagnetic interference capability of the receiver in other drones.

[0046] In this embodiment, after the target drone is equipped with a PTZ lamp and a camera, the initial action video of the circuit breaker opening and closing is captured, and the initial action video is transmitted to the collector of the target drone, and then transmitted to the image processing module in the host computer, and then received from the initial action video stored in the image processing module, wherein the target drone may be a rotary-wing drone. The initial action video may be referred to as the captured video.

[0047] For example, when obtaining the action video of the circuit breaker opening and closing, a multi-rotor drone equipped with a gimbal light and a camera is used to shoot the action video of the circuit breaker operating mechanism controlling the circuit breaker, and the obtained video is transmitted to the drone collector and then to the image processing module in the host computer.

[0048] Optionally, the image processing module may include: a video frame supplement unit, an image calibration unit, an image enhancement unit and a target tracking unit, which are used to calculate various parameters of the circuit breaker opening and closing speed.

[0049] In this embodiment, in a special environment such as a substation, due to the presence of a large number of high-voltage lines and equipment, strong electromagnetic interference will be generated, which will in turn interfere with the stability of data link transmission. Therefore, a target drone is designed to solve the problem of unstable remote transmission of drone data links in complex environments and improve the target drone's anti-electromagnetic interference capability, wherein the target drone is a drone that has been subjected to anti-electromagnetic interference processing.

[0050] Optionally, after the UAV is subjected to anti-electromagnetic interference processing, the process of obtaining the target UAV is as follows: a perception module is designed for the UAV, and when the data terminal in the target UAV transmits inspection data (i.e., the UAV data link), the monitoring platform of the perception module in the anti-electromagnetic interference subsystem can autonomously sense the interference level. When the interference signal breaks through the warning line, the adaptive response module in the target UAV controls the ground transmission power to ensure that the target UAV transmits the inspection data back to the ground station in real time and completely, thereby achieving the purpose of anti-electromagnetic interference; when the monitoring platform senses that the onboard receiver of the inspection subsystem is subjected to electromagnetic interference, the wavelet threshold noise reduction method is used to remove the high-frequency alternating electromagnetic noise therein, and an equivalent current magnetic disturbance compensation model is built to achieve the goal of anti-electromagnetic interference.

[0051] Optionally, when the interference signal exceeds the warning line, the adaptive response module controls the ground transmission power to ensure that the target UAV transmits the inspection data to the ground station in real time and completely. The process is as follows: the UAV data link electromagnetic interference level is used to control the ground transmission power. It can be divided into four levels, namely, locked A, unstable B, about to lose lock C, and lost lock D. During the inspection operation, the UAV detects and obtains the interference signal through the airborne electromagnetic interference environment monitoring system. Suppose R SNs is the signal-to-noise ratio, f j is the interference frequency of signal j, R SNj is the interference signal noise ratio, then according to the above data, the detection input sample can be obtained as {R SNs , f j}. ISj Description SNj The corresponding level D effect interference signal-to-noise ratio can be used to obtain the unlocking effect threshold S j =R ISj P S =P n R SNs R ISj , where P S =P n R SNs The power of the interference signal detected by the monitoring platform can be described. The frequency of the nth signal detected can be expressed by P n The predicted value obtained by the designed training model can be compared with the interference signal, and the electromagnetic interference margin Δ can be solved using the following formula:

[0052] Δ=S j -P s =P n R SNs R ISj -P n R SNs

[0053] Among them, when Δ is not greater than zero, the state of the drone data link is level D; when Δ is greater than zero, the state of the drone data link may be one of the three states of A / B / C. When the drone data link is subjected to co-frequency electromagnetic interference, with the enhancement of the interference signal, its interference change process is more refined and prominent. When in state C, the fluctuation amplitude of the interference signal intensity is maintained at 2 decibels (dB), and it can reach 4dB continuously when in state B. When the interference margin is in the interference warning interval and 0<Δ<2, the interference level of the drone at this time can be classified as state C; if 2≤Δ≤6, the interference level of the drone at this time can be classified as state B; if Δ>6, it is classified as state A.

[0054] Furthermore, since the working signal is very likely to be refracted during propagation, an attenuation model (Longley-Rice model) is used to predict the attenuation of the working signal. The advantage of this model is its strong inclusiveness and wide applicability. According to the inspection distance of the UAV, the attenuation coefficient E can be derived and solved, and the ground transmission power P can be solved by the following formula: t :P t =P S +E-(G r +G t ), where G r Used to represent the gain of the receiving antenna, G t Used to indicate the gain of the transmitting antenna.

[0055] In this embodiment, when the monitoring platform senses that the drone's airborne receiver is interfered by current electromagnetic noise, it is divided into high-frequency alternating electromagnetic noise and equivalent current electromagnetic noise. Among them, the high-frequency alternating electromagnetic noise can be removed by wavelet threshold noise reduction method, and the equivalent current electromagnetic noise can be based on the noise energy distribution ridge line, and the magnetic disturbance compensation model can be built to achieve the purpose of anti-electromagnetic interference.

[0056] Optionally, the process of removing the high-frequency alternating electromagnetic noise by using the wavelet threshold denoising method is as follows: using the wavelet decomposition technology, selecting appropriate wavelet basis functions and decomposition scales, decomposing the high-frequency alternating electromagnetic noise signal, and the high-frequency alternating electromagnetic noise signal can be represented by B. The high-frequency alternating electromagnetic noise signal B is concentrated in the high-frequency first-layer wavelet coefficients after decomposition. Therefore, a three-layer wavelet decomposition can be performed at this time, and the hard threshold method is selected to remove the first-layer high-frequency wavelet coefficients, and the soft threshold method can be used after the second layer to achieve the initial denoising and smoothing goals of the original electromagnetic interference noise signal. Then, the signal set F after the noise is removed is e It can be expressed by the following formula: F e ={f e 1 , f e 2 , fe 3 , ..., f e M}, where M is used to represent the number of signals in the signal set after wavelet decomposition and threshold processing; f e M Used to represent the Mth signal in the signal set after wavelet decomposition and threshold processing.

[0057] Optionally, the process of extracting the high-frequency alternating magnetic disturbance energy ridge is as follows: after the initial noise reduction is completed, the electromagnetic noise is roughly removed, and the high-frequency noise in the previous link is analyzed by using the time-frequency analysis method to obtain the noise energy dispersion status, and according to the characteristics of the high-frequency alternating magnetic disturbance energy scattered in the time domain, the energy ridge shape X is extracted from it, and the energy ridge shape X can be expressed by the following formula: X = (x1, x2, x3, ..., x M ), and regard its corresponding equivalent current electromagnetic noise as the compensation parameter of the equivalent current electromagnetic noise, where x M Used to represent noise energy.

[0058] Optionally, the process of building an equivalent current magnetic disturbance compensation model is as follows: when the UAV system is stationary, the monitoring platform detects that the electromagnetic interference signal changes from low-speed uniform acceleration to high-speed magnetic disturbance. Theoretically, when there is no electromagnetic interference and the detection result is a constant magnetic field, the intensity of electromagnetic interference to the UAV system can be linearly related to the equivalent current intensity. Assume that the equivalent current magnetic noise intensity B e , can be obtained by formula B e =aX+b, where a and b are used to represent the compensation coefficient. The actual electromagnetic background field is expressed by the formula F={f1,f2,f3,…,f M}, where f M is x M The background field strength can be calculated by using a univariate linear regression model to construct an equivalent current magnetic disturbance compensation model, which can be expressed by the following formula:

[0059]

[0060] In the above formula, the compensation coefficients a and b are usually between 0 and 1. If the compensation coefficient is greater than 95%, F e With B e Add together to compensate for the signal loss between the electromagnetic and monitoring platform; otherwise, F e With B e Subtraction removes noise.

[0061] It should be noted that when the monitoring platform senses that the drone's onboard receiver is interfered by current electromagnetic noise, it divides it into high-frequency alternating electromagnetic noise and equivalent current electromagnetic noise, and then cycles the above three processes to clear the electromagnetic noise in the signal received by the receiver and compensate for the interference signal after noise reduction until the onboard receiver achieves electromagnetic interference resistance and stops the cycle.

[0062] As an optional implementation method, image enhancement processing is performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the multiple target images are used to generate a target action video of circuit breaker opening and closing, including: frame supplementation processing is performed on multiple initial images in the initial action video respectively to obtain multiple first images, and the multiple first images are used to generate a first action video of circuit breaker opening and closing, wherein the playback smoothness of the first action video is higher than the playback smoothness of the initial action video, and the multiple first images correspond one-to-one to the multiple initial images; de-distortion processing is performed on multiple first images in the first action video respectively to obtain multiple second images, and the multiple second images are used to generate a second action video of circuit breaker opening and closing, wherein the distortion degree of the second image is lower than the distortion degree of the first image, and the multiple second images correspond one-to-one to the multiple first images; image enhancement processing is performed on multiple second images in the second action video respectively to obtain the target action video.

[0063] In this embodiment, the video frame interpolation unit in the image processing module can be used to interpolate multiple initial images in the initial action video to obtain multiple first images, and the multiple first images can be used to generate a first action video of the circuit breaker opening and closing. Then, the image calibration unit in the image processing module can be used to dedistort the multiple first images in the first action video to obtain multiple second images, and the multiple second images can be used to generate a second action video of the circuit breaker opening and closing. Finally, the image enhancement unit in the image processing module can be used to enhance the multiple second images in the second action video to achieve the purpose of obtaining the target action video.

[0064] Optionally, a trained real-time interpolation video enhancement (RIFE) algorithm is used in the video interpolation unit to interpolate the action image of the circuit breaker operating mechanism, and the processing process is as follows: the RIFE algorithm uses a neural network (IFNet network) to generate an intermediate frame. First, two adjacent frames are input into the IFNet network, and the deformation map from the previous frame (0 time) to the intermediate frame is calculated using the following formula:

[0065]

[0066] In the above formula, F t→0 It is used to characterize the optical flow from the previous frame (time 0) to the intermediate frame. Then, the deformation map from the next frame (time 1) to the intermediate frame is calculated using the following formula:

[0067]

[0068] In the above formula, F t→1 Used to characterize the optical flow from the next frame (1 moment) to the intermediate frame.

[0069] Furthermore, a preliminary frame is constructed according to the following formula, that is, the deformation map obtained by the previous frame is fused with the deformation map obtained by the next frame:

[0070]

[0071] In the above formula, It is used to represent the fusion image obtained by fusing two deformation images. Finally, the intermediate frame can be obtained by lifting the preliminary frame according to the following formula:

[0072]

[0073] Among them, Δ is used to represent the fused residual image, which can improve the details of the intermediate frame. The frame is supplemented from the starting moment when the operating mechanism in the circuit breaker performs the action. Since the circuit breaker moves quickly, it is best to loop the frame supplement 10 times. At the maximum moment of the circuit breaker action, the frame supplement loop should be more than 15 times.

[0074] In this embodiment, after the video frame interpolation unit interpolates multiple images in the initial action video, a first action video is obtained, and then the first action video needs to be dedistorted by the image calibration unit to obtain multiple second images. The multiple second images can form a second action video, where the second images can be called corrected images.

[0075] Optionally, after the video frame interpolation unit completes the frame interpolation, the completed image video is transmitted to the image calibration unit for de-distortion processing. The action video image of the circuit breaker operating mechanism is de-distorted based on the preset distortion coefficient to obtain a corrected image, wherein the preset distortion coefficient is to calibrate the camera of the drone through the Zhang calibration method, and the specific process is: print a black and white chessboard with accurate side length; shoot several chessboards at different positions or angles; detect the corner points of the chessboard; use the camera imaging principle and the relationship between the corner points of the chessboard to calibrate the camera's internal parameter parameters, radial distortion parameters and tangential distortion parameters.

[0076] It should be noted that each calibrated image will provide an external parameter matrix to represent the conversion relationship from the world coordinate system to the camera coordinate system, and the image will be corrected and dedistorted according to the calibration results. The line dancing image taken by the drone is decomposed into the original image, and the original image is transformed from the pixel coordinate system to the camera coordinate system. In the camera coordinate system, the image is dedistorted according to the calibrated intrinsic parameters, radial distortion parameters and tangential distortion parameters, and then the dedistorted image is converted from the camera coordinate system back to the pixel coordinate system as a new image, so as to achieve the purpose of obtaining the corrected image.

[0077] In addition, ideally, the object point P(X W ,Y W ,Z W ) can be projected onto the imaging plane, and the position can be expressed as Because of the distortion of the image, the actual phase point will move to the position The aberration (δ x ,δ y ):

[0078]

[0079] Optionally, according to the distortion effect generated, the distortion is divided into radial distortion and tangential distortion. The radial distortion coefficient can be obtained by using a common radial distortion model, and the common radial distortion model can be expressed by the following formula:

[0080]

[0081] In the above formula, (δ kx ,δ ky ) is used to represent the radial distortion coefficient, (k1, k2, k3) is used to represent the tangential distortion coefficient, The tangential distortion coefficient is obtained using the commonly used tangential distortion model. The commonly used tangential distortion model can be expressed by the following formula:

[0082]

[0083] In the above formula, (δ px ,δ py ) is used to represent the tangential distortion coefficient, and (p1,p2) is used to represent the radial distortion coefficient.

[0084] The actual imaging model of the camera can be obtained from the above formula, and the actual imaging model can be expressed by the following formula:

[0085]

[0086] Therefore, the actual imaging position of the object point P can be obtained, which is also the actual imaging model of the camera, and the correction image can be obtained by performing distortion correction on each pixel point in the action image of the operating mechanism. From the correction image, the second action video, that is, the correction video, can be determined, and then processed by the image enhancement unit to achieve the purpose of obtaining the target action video.

[0087] As an optional implementation method, image enhancement processing is performed on multiple second images in the second action video to obtain a target action video, including: based on the detection requirement information of the circuit in which the circuit breaker is located, multiple weight values ​​corresponding to the multiple second images are determined; based on the multiple weight values, grayscale processing is performed on the multiple second images to obtain multiple third images, wherein the multiple third images correspond one-to-one to the multiple second images; image enhancement processing is performed on the multiple third images to obtain the target action video.

[0088] In this embodiment, after obtaining the second action video, multiple weight values ​​corresponding to multiple second images can be determined respectively according to the detection requirement information of the circuit in which the circuit breaker is located, and then according to the multiple weight values ​​obtained above, the image enhancement unit is called, and the multiple second images are gray-processed respectively using the weighted average method to obtain multiple third images, and then the multiple third images are image-enhanced respectively to obtain the target action video. Among them, the multiple weight values ​​are used to characterize the weight coefficients of the three channels in the image. The third image can be called a grayscale correction image.

[0089] For example, after obtaining the corrected images, the corrected images are combined into a high-speed image sequence, and the corrected images are converted into grayscale corrected images to obtain a high-speed grayscale corrected image sequence, which is then processed and analyzed using a grayscale algorithm.

[0090] For another example, the specific process of converting the correction image into the grayscale correction image is as follows: Since the pixel value of the photo collected by the camera is generally large, and based on the three-channel information, the image will contain more information that is irrelevant to the opening and closing stroke curve and the opening and closing speed curve, directly using the three-channel original image for image processing and feature extraction will result in a huge amount of calculation and affect the calculation speed. Therefore, it is necessary to use grayscale transformation, through a certain mapping relationship, based on the three-channel color of each pixel itself, to map it to a single-channel grayscale information, so that the speed of subsequent calculations can be improved while retaining the execution of the circuit breaker's operating mechanism.

[0091] It should be noted that the grayscale of the correction image can be transformed by using the weighted average method. According to the requirements of the power transmission line dancing detection, different weight coefficients are assigned to the brightness of the three channels (R, G, B) to obtain a grayscale correction image. The grayscale correction image can be expressed by the following formula:

[0092]

[0093] The weight value is determined according to the sensitivity of the human eye to color. The final weight coefficients of the three channels can be w r =0.229,w g =0.589,w b = 0.114. The weight coefficients obtained above are used to realize image transformation and obtain a grayscale corrected image.

[0094] As an optional implementation method, image enhancement processing is performed on multiple third images respectively to obtain target action videos, including: performing histogram equalization processing on multiple third images respectively to obtain multiple fourth images, wherein the contrast of the fourth images is higher than the contrast of the third images, and the multiple fourth images correspond one-to-one to the multiple third images; performing median filtering processing on the multiple fourth images respectively to obtain multiple target images, wherein the overall image quality of the target images is higher than the overall image quality of the fourth images, and the multiple target images correspond one-to-one to the multiple fourth images.

[0095] In this embodiment, the multiple third images can be subjected to histogram equalization processing to obtain multiple fourth images, and then the multiple fourth images can be subjected to median filtering processing to obtain multiple target images. The histogram equalization processing can be implemented by an improved histogram equalization algorithm. The median filtering processing can be implemented by a median filtering algorithm.

[0096] Optionally, when the drone collects the action image of the operating mechanism, it is often affected by the illumination, and the illumination on the action image is uneven, and some areas are too bright or too dark, which will affect the subsequent image recognition operation. Therefore, image enhancement is needed to improve the quality of the grayscale correction image. After the histogram method is adopted, the spacing between each grayscale level becomes wider and flatter. Moreover, the pixels are not concentrated in a certain section, and the overall is more uniform, which improves the distinction between each grayscale level, which is beneficial to the next step of image analysis. However, at the same time, after histogram equalization, different pixel levels may become the same, forming the same grayscale value, and there will be obvious distinctions between different grayscale value areas, and even pseudo contours and over-enhancement phenomena will occur. Therefore, the grayscale correction image is enhanced by the improved histogram equalization algorithm to achieve an image that can retain the main edge information of the transmission line without generating pseudo contours.

[0097] Among them, the improved histogram equalization algorithm can be expressed by the following formula:

[0098]

[0099] In the above formula, n kUsed to indicate the number of occurrences of the kth gray value, r k is used to represent the kth grayscale value, L is used to represent the number of grayscale levels, and N is used to represent the total number of pixels P r (r k ) is r k The probability of occurrence.

[0100] Furthermore, in the process of histogram equalization, no matter how the pixels are mapped, the original size relationship must be kept unchanged, that is, the brighter area is still brighter, and the darker area is still darker, but the contrast is increased, and the brightness and darkness must not be reversed. The specific steps of histogram equalization are: count the number of pixels n at each gray level in the grayscale correction image k ; Use the above formula to calculate the probability P of each gray level appearing in the grayscale correction image r (r k ); Calculate the cumulative histogram of different gray levels, wherein the process of calculating the cumulative histogram can be expressed by the following formula:

[0101]

[0102] According to the above formula, the histogram equalization formula can be used to solve the gray level g of the transformed image. i , then the gray level g i It can be expressed by the following formula:

[0103] g i = int[(g max -g min ) k +0.5)

[0104] Where i = 1, 2, ..., P-1, P is used to represent the number of gray levels of the transformed image, g max Used to represent the maximum gray value in the grayscale correction image, g min Used to represent the minimum grayscale value; then get the coordinate pixel value mapped back to the original grayscale correction image; use the following formula to calculate the proportion of each grayscale pixel to the total pixels P g (g i ):

[0105]

[0106] In the embodiment of the present invention, after adopting the improved histogram equalization algorithm, a fourth image f can be obtained, which can be expressed by the formula f=I γ =(g i ) γThe main edge information of the operating mechanism of the circuit breaker is retained without generating pseudo contours, maximizing the difference between the target and the background, and significantly enhancing the contrast between the operating mechanism and the background. At this time, since most circuit breakers will introduce certain noise during the opening and closing process, it is also necessary to denoise the obtained grayscale correction image.

[0107] Optionally, the process of denoising the grayscale corrected image after histogram equalization is: using median filtering to denoise the grayscale corrected image after histogram equalization, wherein the median filtering is to filter the image without losing the image edge as much as possible, thereby improving the quality of the grayscale corrected image.

[0108] Optionally, after median filtering, the pixel of each point is the median value in the neighborhood of this pixel, that is, the pixels in the template area are sorted according to the size of the pixel value, and then each pixel value takes the middle value of the template sorting centered on this point, and the template size is generally 3*3 or 5*5. The template sliding of the median filter is similar to the convolution calculation, which can exclude relatively isolated noise points and is not sensitive to the foreground or background part of the image, so it can retain more image details while suppressing noise.

[0109] Optionally, the filtering process is as follows: place the filter template in the upper left corner of the grayscale correction image; each time the template is fixed, all the grayscale values ​​in the template are stored, and then these pixel values ​​are sorted from large to small; the number of pixels in the template is an odd number, so the middle value in the pixel sorting is placed in the pixel where the center of the template is located; after the calculation is completed, move the template, and after each pixel is calculated, move one grid to the right, and after one row is completed, proceed to the next row in turn until the entire image is traversed. During the filtering process, if the template moves to the row and column at the boundary, for example, when the center of the template starts to move from the upper left corner, due to the symmetry of the template, some templates will have no values. At this time, you only need to sort the existing values ​​and find the median, or copy the values ​​of the pixels next to it that can be fully calculated.

[0110] Optionally, the process of median filtering can be expressed by the following formula:

[0111] d(x,y)=Mid{p(xm,yn),(m,n∈W)}

[0112] Among them, p(x,y) is used to represent the pixel value of each point in the original image, and d(x,y) is used to represent the pixel value of the point after median filtering.

[0113] As an optional implementation method, target tracking processing is performed on the target action video to generate a running trajectory of the circuit breaker in the target action video, including: performing corner point detection on multiple target images in the target action video to obtain target corner points of the multiple target images; inputting the target corner points into a target tracking processing model for analysis to generate the running trajectory of the circuit breaker, wherein the target tracking processing model is used to be established through target constraint conditions, and the target constraint conditions are used to maintain the continuity of the target action video.

[0114] In this embodiment, a target tracking unit is used to detect corner points of multiple target images in the obtained target action video to obtain target corner points of multiple target images, and then the obtained target corner points are input into the target tracking processing model for analysis to generate the running trajectory of the circuit breaker. The target action video is used to represent the enhanced video.

[0115] Optionally, after the initial action video is processed by the image enhancement unit to obtain the target action video, the enhanced video is transmitted to the target tracking unit to obtain the operation trajectory of the circuit breaker.

[0116] Optionally, corner detection is a detection method for obtaining image features in visual processing technology. It is the first step in parsing video images. It is widely used in moving target tracking, image registration, target pose estimation, etc. It can also be called feature point detection. It is mainly divided into global detection and local area detection. Global detection requires analysis of the entire image and is suitable for mobile camera shooting or car tracking.

[0117] Optionally, the high-speed camera is stationary when filming the execution action of the operating mechanism of the high-voltage circuit breaker, so local area detection can be used, and the detection area is the dot area. This detection method determines whether there is a corner point based on the change in the grayscale value of the local pixel point, and the amount of calculation of this calculation method is greatly reduced compared to the global detection, the calculation speed is fast, simple and efficient. Among them, another corner point detection algorithm (Shi-Tomasi algorithm) is a corner point discrimination method based on a window sliding model. This method can determine whether a feature point is a corner point by whether the first-order derivative of the grayscale is the largest within a certain range. If the center point of the window is I (x, y) and the translation is (u, v), the grayscale change of the window can be expressed by the following formula:

[0118] F(x,y)=Σ (x,y)∈S ω(x,y)[I(x+u,y+v)-I(x,y)] 2

[0119] In the above formula, S is used to represent the area where the window moves, w(x,y) is used to represent the window function, and different weights can be assigned according to different needs. I(x,y) is used to represent the image grayscale, and I(x+u,y+v) is used to represent the grayscale of the image after translation.

[0120] From I(x+u,y+v)=I(x,y)+uI x +vI y +O(u 2 ,v 2 ) to obtain the following formula:

[0121]

[0122] For a small local movement [u,v], it can be expressed by the following formula:

[0123]

[0124] Among them, M is a 2*2 matrix, which can be obtained by the following formula:

[0125]

[0126] According to the above formula, the metric corner function can be obtained:

[0127] R = λ1λ2-k(λ1+λ 21 ) 2

[0128] In the above formula, R is the corner response function, and λ1 and λ1 are the characteristic values ​​of the pixel gradient. Threshold processing is performed on the function R:

[0129] R>threshold

[0130] Among them, threshold is the set threshold. When the two λ1 and λ2 of (x, y) are input into the formula for obtaining R, if the output R is greater than the threshold, the point (x, y) is determined to be a corner point. After the corner point is detected, the tracking algorithm is used to track the corner point.

[0131] As an optional implementation method, the target constraint condition includes: a first constraint condition and a second constraint condition, wherein the first constraint condition is used to constrain the image brightness between adjacent frames in multiple target images respectively, and the second constraint condition is used to constrain the image conversion time between adjacent frames.

[0132] In this embodiment, the target constraint conditions include: a first constraint condition, a second constraint condition and a third constraint condition, wherein the third constraint condition is used to constrain the positions of adjacent pixels in multiple target images respectively, and the first constraint condition, the second constraint condition and the third constraint condition are three assumptions based on the optical flow method respectively.

[0133] Optionally, the target tracking processing model uses an improved KLT optical flow method with front and rear threshold conditions for calculation, extracting and matching feature points, and using the sum of squared grayscale differences of a window of video information in a specific image as a metric to complete the algorithm operation.

[0134] Optionally, the first constraint condition can be used to characterize and constrain the brightness of image pixels between adjacent frames to be constant. The second constraint condition can be used to ensure that the time between adjacent frames is short enough or the object movement is small; and the third constraint condition is used to ensure that adjacent pixels remain adjacent in the next frame.

[0135] For example, according to the first constraint of the optical flow method, the action video image of the circuit breaker operating mechanism contains a feature with the same grayscale value at the tth time and the t+dtth time. Assume that the pixel coordinates of the grayscale value I(x, y, t) of this feature at the tth time are (x, y), and the pixel coordinates of the grayscale value I(x+dx, y+dy, t+dt) at the t+dtth time are (x+dx, y+dy), then the grayscale value I(x, y, t) can be expressed by the following formula:

[0136] I(x,y,y)=I(x+dx,y+dy,t+dt)

[0137] Performing a first-order Taylor expansion on the right side of the above equation gives the following formula:

[0138]

[0139] Based on the above formula, let d = [dx, dy] be the offset of the pixel point, that is, the optical flow of the feature point, and the size of the window W between the two frames is (2w x +1)×(2w y +1), the center point is (c x ,c y ). The feature point matching process of the KLT optical flow method is to find the offset d under the feature window W so that its residual function ε(d) is minimized. The pixels and actual values ​​of the image can be expressed by the following formula:

[0140]

[0141] Based on the above formula, let the derivative of the extreme point be zero, and find the minimum value of the residual function ε(d):

[0142]

[0143] in, Used to represent spatial gradients of video images.

[0144] The formula Substitute into the formula The following formula can be obtained:

[0145]

[0146] make Can get

[0147] d=G -1 ·e

[0148] In the above formula, d is the displacement, and the target can be tracked by iteratively calculating each corner point in the algorithm.

[0149] In this embodiment, based on the second constraint condition of the optical flow method, it can be seen that the KLT optical flow method is suitable for objects with small motion between two frames. However, in fact, the opening and closing speed of the high-voltage circuit breaker is relatively fast, and the displacement is large near the maximum speed, which may cause target drift and mismatch. Therefore, a comparison step of the front and rear thresholds is added in the target tracking process to improve the robustness of tracking.

[0150] Optionally, assume that the feature points of the n-1th frame are estimated by the KLT optical flow method to move the distance of the nth frame through d f(n-1) Represented, and then inversely calculated from the nth frame to the n-1th frame, and the stroke can be obtained by d b(n-1) In theory, d f(n-1) =d b(n-1) , which is used to indicate that the feature point positions on the two frames coincide, but in fact, the optical flow estimation of the circuit breaker near the maximum speed will cause the positions of some feature points to shift. The offset value can be calculated by the following formula:

[0151] Δd=|d f(n-1) -d b(n-1) |<T1

[0152] In the above formula, if Δd<T1, it means that the feature point can be reliably tracked. On the contrary, if Δd>T1, it means that the feature point cannot be stably tracked and needs to be deleted. After many experiments, it was found that setting T1 to 0.8 is the most appropriate. In the video image of the circuit breaker's opening and closing action, adjacent pixels remain adjacent in the next frame, which meets the third constraint of the optical flow method.

[0153] In this embodiment, an initial action video of the circuit breaker opening and closing can be obtained first, and then image enhancement processing can be performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the multiple target images are used to generate a target action video of the circuit breaker opening and closing. Then, target tracking processing is performed on the target action video to generate an operation trajectory of the circuit breaker in the target action video. Finally, according to the operation trajectory obtained above, the purpose of determining the target speed of the circuit breaker opening and closing is achieved, thereby solving the technical problem of difficulty in obtaining a suitable circuit breaker opening and closing speed and achieving the technical effect of obtaining a suitable circuit breaker opening and closing speed.

[0154] The technical solution of the embodiment of the present invention is illustrated below in conjunction with preferred implementation modes.

[0155] At present, the opening and closing speed of circuit breakers is an important indicator for evaluating the reliability of circuit breakers. The commonly used method to obtain the opening and closing speed of circuit breakers is to use special testing instruments, such as opening and closing speed testers or oscilloscopes, to measure the opening and closing time of circuit breakers to determine whether their speed meets the standard requirements. If the opening and closing speed of the circuit breaker obtained is too fast, it may cause excessive impact on the operating mechanism, thereby damaging the components; if it is too slow, it may cause the arcing time between the contacts to be prolonged, increasing electrical wear. As a result, it is difficult to obtain a suitable opening and closing speed of the circuit breaker.

[0156] In traditional maintenance work, a common method is to install a sliding resistance displacement sensor or an acceleration sensor on a special detection tool, and then fix the tool on the transmission connecting rod of the circuit breaker. When the connecting rod moves, the sensor can capture the movement of the connecting rod, thereby indirectly knowing the speed characteristics of the circuit breaker. However, the above method still has the technical problem of difficulty in obtaining the appropriate circuit breaker opening and closing speed.

[0157] Therefore, in order to solve the above problems, the present invention proposes a circuit breaker opening and closing speed detection system based on drone image acquisition, which combines image recognition and processing technology, adopts a non-contact rapid detection method, and remotely obtains the speed data of circuit breaker opening and closing. The system mainly includes two modules: an image data acquisition module and an image data processing module. Among them, the image data acquisition module includes a drone, a drone control system, a perception module, and a fill light unit. This module collects the action image of the circuit breaker opening and closing and transmits it to the host computer; the image data processing module includes a video frame supplement unit, an image calibration unit, an image enhancement unit, and a target tracking unit, which are used to calculate the various parameters of the circuit breaker opening and closing speed. The application of this speed detection system solves the technical problem of difficulty in obtaining a suitable circuit breaker opening and closing speed, and achieves the technical effect of being able to obtain a suitable circuit breaker opening and closing speed.

[0158] In an embodiment of the present invention, Figure 2The present invention is a flow chart of a circuit breaker opening and closing speed detection system based on drone image acquisition according to an embodiment of the present invention, and the flow chart includes the following steps:

[0159] In step S201, the host computer sends a collection instruction to the drone collection controller, starts the drone gimbal light, and starts shooting the dot mark on the operating mechanism. After the high-voltage circuit breaker completes the opening and closing operation, it turns off the shooting and transmits the video wirelessly to the image processing module in the host computer.

[0160] In this embodiment, when the circuit breaker performs opening and closing actions, a multi-rotor drone equipped with a gimbal light and a camera is used to capture an action video of the circuit breaker operating mechanism controlling the circuit breaker, and the captured video is transmitted to the drone collector and then to the image processing module in the host computer.

[0161] In this embodiment, Figure 3 is a schematic diagram of a drone system according to an embodiment of the present invention. Figure 3 As shown, the schematic diagram of the UAV system includes: a duplexer 301, a power divider 302, a receiver 303, a transmitter 304, a monitoring platform 305, a memory 306, a data terminal 307 and a flight control module 308, wherein the monitoring platform 305 includes a receiver 3051 and a UAV data link 3052.

[0162] Optionally, the data terminal can input the acquired relevant data into the transmitter, and the transmitter processes the above-obtained relevant data, and inputs the processed data and the electromagnetic interference signal into the duplexer. After being processed by the duplexer, it can be input into the power divider for further processing, and then the data processed by the power divider is input into the monitoring platform, so as to improve the anti-electromagnetic interference capability of the receiver in the UAV. The memory can be used to train and predict the acquired data, and then after adaptive response, a response instruction can be obtained, and the response instruction is sent to the flight control module to obtain telemetry instructions and shooting instructions, and the shooting instructions are sent to the monitoring platform, and the telemetry instructions are sent to the data terminal, and the data terminal can send the relevant information of the telemetry feedback to the flight control module. Among them, the monitoring platform includes: a receiver and a UAV data link.

[0163] In this embodiment, in a special environment such as a substation, due to the presence of a large number of high-voltage lines and equipment, strong electromagnetic interference will be generated, which will in turn interfere with the stability of data link transmission. Therefore, a target drone is designed to solve the problem of unstable remote transmission of drone data links in complex environments and improve the target drone's ability to resist electromagnetic interference.

[0164] Optionally, the core of the anti-interference design is the perception module. When the data terminal in the target UAV transmits inspection data (i.e., the UAV data link), the monitoring platform of the perception module in the anti-electromagnetic interference subsystem can autonomously sense the interference level. When the interference signal exceeds the warning line, the adaptive response module in the target UAV controls the ground transmission power to ensure that the target UAV transmits the inspection data to the ground station in real time and completely, thereby achieving the purpose of anti-electromagnetic interference. When the monitoring platform senses that the onboard receiver of the inspection subsystem is subject to electromagnetic interference, the wavelet threshold noise reduction method is used to eliminate the high-frequency alternating electromagnetic noise, and an equivalent current magnetic disturbance compensation model is built to achieve the goal of anti-electromagnetic interference.

[0165] In this embodiment, after the electromagnetic interference level of the drone data link is determined by the adaptive response module in the subsystem, the ground transmission power is controlled, and the inspection data is transmitted back to the ground station via the downlink telemetry link to enhance the signal strength received by the drone and achieve the goal of anti-electromagnetic interference. When the monitoring platform senses that the drone's airborne receiver is interfered by current electromagnetic noise, it is divided into high-frequency alternating electromagnetic noise and equivalent current electromagnetic noise. Among them, the high-frequency alternating electromagnetic noise can be removed by wavelet threshold noise reduction method, and the equivalent current electromagnetic noise can be based on the noise energy distribution ridge to build a magnetic disturbance compensation model to achieve the purpose of anti-electromagnetic interference.

[0166] Optionally, the process of removing high-frequency alternating electromagnetic noise is: using wavelet decomposition technology, selecting appropriate wavelet basis functions and decomposition scales, decomposing high-frequency alternating electromagnetic noise signals, and the high-frequency alternating electromagnetic noise signals can be represented by B. The high-frequency alternating electromagnetic noise signal B is concentrated in the high-frequency first-layer wavelet coefficients after decomposition. Therefore, a three-layer wavelet decomposition can be performed at this time, and a hard threshold method is selected to remove the first-layer high-frequency wavelet coefficients, and a soft threshold method can be used after the second layer to achieve the initial noise reduction and smoothing goals of the original electromagnetic interference noise signal. The signal set F after the noise is removed is e It can be expressed by the following formula: F e ={f e 1 , f e 2 , f e 3 , ..., f e M}, where M is used to represent the number of signals in the signal set after wavelet decomposition and threshold processing; f e M Used to represent the Mth signal in the signal set after wavelet decomposition and threshold processing.

[0167] Optionally, the process of extracting the high-frequency alternating magnetic disturbance energy ridge is as follows: after the initial noise reduction is completed, the electromagnetic noise is roughly removed, and the high-frequency noise in the previous link is analyzed by using the time-frequency analysis method to obtain the noise energy dispersion status, and according to the characteristics of the high-frequency alternating magnetic disturbance energy scattered in the time domain, the energy ridge shape X is extracted from it, and the energy ridge shape X can be expressed by the following formula: X = (x1, x2, x3, ..., x M ), and regard its corresponding equivalent current electromagnetic noise as the compensation parameter of the equivalent current electromagnetic noise, where x M Used to represent noise energy.

[0168] Optionally, the process of building an equivalent current magnetic disturbance compensation model is as follows: when the UAV system is stationary, the monitoring platform detects that the electromagnetic interference signal changes from low-speed uniform acceleration to high-speed magnetic disturbance. Theoretically, when there is no electromagnetic interference and the detection result is a constant magnetic field, the intensity of electromagnetic interference to the UAV system can be linearly related to the equivalent current intensity. Assume that the equivalent current magnetic noise intensity B e , can be obtained by formula B e =aX+b, where a and b are used to represent the compensation coefficient. The actual electromagnetic background field is expressed by the formula F={f1,f2,f3,…,f M}, where f M is x M The background field strength can be calculated by using a univariate linear regression model to construct an equivalent current magnetic disturbance compensation model.

[0169] Optionally, when the monitoring platform senses that the drone's onboard receiver is subject to current electromagnetic noise interference, it divides it into high-frequency alternating electromagnetic noise and equivalent current electromagnetic noise, and then cycles the above three processes to clear the electromagnetic noise in the signal received by the receiver and compensate for the interference signal after noise reduction until the onboard receiver achieves electromagnetic interference resistance and stops the cycle.

[0170] Step S202, in the image processing module, the video interpolation unit first interpolates the image taken by the drone, then the image calibration unit dedistorts the image, and then the enhancement unit performs grayscale transformation, histogram equalization, median filtering and other processing on the image. Finally, the target tracking unit performs corner point detection and adopts the improved KLT optical flow method to track the target, and obtains the travel curve of the high-voltage circuit breaker.

[0171] In this embodiment, an image processing module is used to perform image enhancement processing on the video taken by the drone, so as to ultimately achieve the purpose of obtaining the travel curve of the high-voltage circuit breaker by using corner point detection and an improved KLT optical flow method.

[0172] Optionally, Figure 4is a schematic diagram of an image processing module according to an embodiment of the present invention, such as Figure 4 As shown, the image processing module includes: a video frame supplement unit 401, an image calibration unit 402, an image enhancement unit 403 and a target tracking unit 404. Since the circuit breaker has a high speed of opening and closing, the acquisition rate of a high-speed camera can reach thousands of frames per second. Generally, the acquisition is performed by tracking the action moment through a high-speed camera. The frame number of the drone is low, and it is impossible to perfectly capture every frame of the circuit breaker operating mechanism action. Therefore, the frame supplement method is used to supplement the frame to improve the target tracking effect.

[0173] In this embodiment, after the video frame interpolation unit completes the frame interpolation, the image video after the frame interpolation is completed is transmitted to the image calibration unit for de-distortion processing. The action video image of the circuit breaker operating mechanism is de-distorted based on the preset distortion coefficient to obtain a corrected image, wherein the preset distortion coefficient is to calibrate the camera of the drone by Zhang's calibration method, and the specific process is: printing a black and white chessboard with accurate side length; shooting several chessboards at different positions or angles; detecting the corner points of the chessboard; using the camera imaging principle and the relationship between the corner points of the chessboard to calibrate the camera's internal parameter parameters, radial distortion parameters and tangential distortion parameters.

[0174] Optionally, each calibrated image will provide an external parameter matrix to represent the conversion relationship from the world coordinate system to the camera coordinate system, and the image will be corrected and dedistorted according to the calibration results. The line dancing image taken by the drone is decomposed into the original image, and the original image is transformed from the pixel coordinate system to the camera coordinate system. In the camera coordinate system, the image is dedistorted according to the calibrated intrinsic parameters, radial distortion parameters and tangential distortion parameters, and then the dedistorted image is converted from the camera coordinate system back to the pixel coordinate system as a new image, thereby achieving the purpose of obtaining the corrected image.

[0175] In this embodiment, based on the corrected video, after determining the corrected video sequence, the image enhancement unit performs processing: the corrected images are combined into a high-speed image sequence, and the corrected images are converted into grayscale corrected images to obtain a high-speed grayscale corrected image sequence, and the high-speed grayscale corrected image sequence is processed and analyzed by a grayscale algorithm.

[0176] For another example, the specific process of converting the correction image into the grayscale correction image is as follows: Since the pixel value of the photo collected by the camera is generally large, and based on the three-channel information, the image will contain more information that is irrelevant to the opening and closing stroke curve and the opening and closing speed curve, directly using the three-channel original image for image processing and feature extraction will result in a huge amount of calculation and affect the calculation speed. Therefore, it is necessary to use grayscale transformation, through a certain mapping relationship, based on the three-channel color of each pixel itself, to map it to a single-channel grayscale information, so that the speed of subsequent calculations can be improved while retaining the execution of the circuit breaker's operating mechanism.

[0177] Then, the grayscale correction image is enhanced by the improved histogram equalization algorithm: when the UAV collects the action image of the operating mechanism, it is often affected by the light, and the action image is prone to uneven illumination. Some areas are too bright or too dark, which will affect the subsequent image recognition operation. Therefore, image enhancement is needed to improve the quality of the grayscale correction image. After the histogram method is adopted, the spacing between each grayscale level becomes wider and flatter. Moreover, the pixels are not concentrated in a certain section, and the overall uniformity is improved, which improves the distinction between each grayscale level, which is conducive to the next step of image analysis. However, at the same time, after histogram equalization, different pixel levels may become the same, forming the same grayscale value. At the same time, there will be obvious distinctions between different grayscale value areas, and even pseudo contours and over-enhancement phenomena will occur. Therefore, the grayscale correction image is enhanced by the improved histogram equalization algorithm to achieve an image that can retain the main edge information of the transmission line without generating pseudo contours.

[0178] In the embodiment of the present invention, after adopting the improved histogram equalization algorithm, a fourth image f can be obtained, which can be expressed by the formula f=I γ =(g i ) γ The main edge information of the operating mechanism of the circuit breaker is retained without generating pseudo contours, maximizing the difference between the target and the background, and significantly enhancing the contrast between the operating mechanism and the background. At this time, since most circuit breakers will introduce certain noise during the opening and closing process, it is also necessary to denoise the obtained grayscale correction image.

[0179] Optionally, the process of denoising the grayscale corrected image after histogram equalization is: using median filtering to denoise the grayscale corrected image after histogram equalization, wherein the median filtering is to filter the image without losing the image edge as much as possible, thereby improving the quality of the grayscale corrected image.

[0180] Optionally, after median filtering, the pixel of each point is the median value in the neighborhood of this pixel, that is, the pixels in the template area are sorted according to the size of the pixel value, and then each pixel value takes the middle value of the template sorting centered on this point, and the template size is generally 3*3 or 5*5. The template sliding of the median filter is similar to the convolution calculation, which can exclude relatively isolated noise points and is not sensitive to the foreground or background part of the image, so it can retain more image details while suppressing noise.

[0181] Optionally, the filtering process is as follows: place the filter template in the upper left corner of the grayscale correction image; each time the template is fixed, all the grayscale values ​​in the template are stored, and then these pixel values ​​are sorted from large to small; the number of pixels in the template is an odd number, so the middle value in the pixel sorting is placed in the pixel where the center of the template is located; after the calculation is completed, move the template, and after each pixel is calculated, move one grid to the right, and after one row is completed, proceed to the next row in turn until the entire image is traversed. During the filtering process, if the template moves to the row and column at the boundary, for example, when the center of the template starts to move from the upper left corner, due to the symmetry of the template, some templates will have no values. At this time, you only need to sort the existing values ​​and find the median, or copy the values ​​of the pixels next to it that can be fully calculated.

[0182] Step S203, tracking is performed according to the tracking corner point, and the travel curve of the high-voltage circuit breaker is obtained by drawing the curve, so as to query the position of the just-opening and closing point, and obtain the opening and closing speed, just-opening and closing speed and maximum speed of the high-voltage circuit breaker.

[0183] In this embodiment, after the image enhancement unit completes processing, the enhanced video image is transmitted to the target tracking unit, and the target tracking unit is used to detect and track the tracking corner points, and a characterization curve is obtained to obtain the travel curve of the high-voltage circuit breaker, thereby achieving the purpose of obtaining the opening and closing speed, the just opening and closing speed and the maximum speed of the high-voltage circuit breaker.

[0184] Optionally, corner detection is a detection method for obtaining image features in visual processing technology. It is the first step in parsing video images. It is widely used in moving target tracking, image registration, target pose estimation, etc. It can also be called feature point detection. It is mainly divided into global detection and local area detection. Global detection requires analysis of the entire image and is suitable for mobile camera shooting or car tracking.

[0185] Optionally, the high-speed camera is stationary when photographing the execution action of the operating mechanism of the high-voltage circuit breaker, so local area detection can be used, and the detection area is the dot area. This detection method determines whether there is a corner point based on the change of the gray value of the local pixel point, and the calculation amount of this calculation method is greatly reduced compared with the global detection, the calculation speed is fast, simple and efficient. Among them, another corner point detection algorithm (Shi-Tomasi algorithm) is a corner point discrimination method based on a window sliding model. This method can determine whether a feature point is a corner point by whether the first-order derivative of the gray value is the largest within a certain range.

[0186] In this embodiment, after the corner point detection, the tracking algorithm is used to track the corner point, and the improved KLT optical flow method based on the front and back threshold conditions is used to extract and match the feature points, and the sum of squared grayscale differences of the window of the video information in a specific image is used as a metric to complete the operation of the algorithm. Among them, the three assumptions of the algorithm based on the optical flow method are as follows: the brightness of the image pixels is constant between adjacent frames; the time between adjacent frames is short enough or the object movement is small; adjacent pixels remain adjacent in the next frame.

[0187] In this embodiment, based on the assumption of the optical flow method that the time between adjacent frames is short enough or the object movement is small, it can be seen that the KLT optical flow method is suitable for the case where the object movement between two frames is small. However, in fact, the opening and closing speed of the high-voltage circuit breaker is fast, and the displacement is large near the maximum speed, which may cause target drift and mismatch. Therefore, a comparison step of the front and rear thresholds is added in the target tracking process to improve the robustness of the tracking.

[0188] Optionally, assume that the feature points of the n-1th frame are estimated by the KLT optical flow method to move the distance of the nth frame through d f(n-1) Represented, and then inversely calculated from the nth frame to the n-1th frame, and the stroke can be obtained by d b(n-1) In theory, d f(n-1) =d b(n-1) , which is used to indicate that the feature point positions on the two frames coincide, but in fact, the optical flow estimation of the circuit breaker near the maximum speed will cause the positions of some feature points to shift. The offset value can be calculated by the following formula:

[0189] Δd=|d f(n-1) -d b(n-1) |<T1

[0190] In the above formula, if Δd<T1, it means that the feature point can be reliably tracked. On the contrary, if Δd>T1, it means that the feature point cannot be stably tracked and needs to be deleted. After many experiments, it was found that setting T1 to 0.8 is the most appropriate. In the video image of the circuit breaker's opening and closing action, adjacent pixels remain adjacent in the next frame, which is consistent with the assumption of the optical flow method that adjacent pixels remain adjacent in the next frame.

[0191] In this embodiment, the corner point is tracked by the improved KLT optical flow method to obtain the corner point motion trajectory, and the distance corresponding to a single pixel in practice is obtained by testing the motion distance of the operating mechanism and the ratio relationship. The corner point motion trajectory is converted into the circuit breaker opening and closing stroke curve, and then according to the definition of the just opening and closing point, the position of the just opening and closing point in the circuit breaker opening and closing stroke curve is determined, and then the circuit breaker opening and closing speed is obtained.

[0192] In an embodiment of the present invention, Figure 5 FIG. 1 is a schematic diagram of obtaining image data of circuit breaker opening and closing collected by a drone according to an embodiment of the present invention. Figure 5 As shown, the schematic diagram mainly includes: a circuit breaker sulfur hexafluoride (SF6) gas chamber 501, a grounding switch 502, a grounding switch control box 503, a circuit breaker operating mechanism 504, a circuit breaker opening and closing control box 505, a drone 506, a drone data collection transmitter 507, a host computer 508 and a coaxial cable 509.

[0193] The circuit breaker sulfur hexafluoride (SF6) gas chamber 501 is used to ensure the normal operation of electrical equipment.

[0194] The grounding switch 502 is used to cut off the grounding fault in the circuit.

[0195] The grounding switch control box 503 is used to control and protect the circuit breaker and the grounding switch.

[0196] In this embodiment, before collecting the action image of the circuit breaker operating mechanism, the circuit breaker opening and closing control box 505 is opened, and a mark is affixed to the circuit breaker operating mechanism 504 inside it; the host of the upper computer 508 issues a collection instruction and transmits it to the drone collection transmitter 507 through the coaxial cable 509, and the drone collection transmitter controls the drone 506 to start aiming and shooting. After the circuit breaker opening and closing is completed, the upper computer issues a closing instruction, the drone stops collecting, and transmits the collected data to the image data processing module on the host.

[0197] In this embodiment, an initial action video of the circuit breaker opening and closing can be obtained first, and then image enhancement processing can be performed on multiple initial images in the initial action video respectively to obtain multiple target images, and the multiple target images are used to generate a target action video of the circuit breaker opening and closing. Then, target tracking processing is performed on the target action video to generate an operation trajectory of the circuit breaker in the target action video. Finally, according to the operation trajectory obtained above, the purpose of determining the target speed of the circuit breaker opening and closing is achieved, thereby solving the technical problem of difficulty in obtaining a suitable circuit breaker opening and closing speed and achieving the technical effect of obtaining a suitable circuit breaker opening and closing speed.

[0198] According to an embodiment of the present invention, a circuit breaker opening and closing speed determination device is provided. It should be noted that the circuit breaker opening and closing speed determination device can be used to execute a circuit breaker opening and closing speed determination method in the embodiment.

[0199] Figure 6 A circuit breaker opening and closing speed determining device according to an embodiment of the present invention is provided. Figure 4 As shown, a circuit breaker opening and closing speed determination device 600 may include: a first acquisition unit 601, a second acquisition unit 602, a generation unit 603 and a determination unit 604.

[0200] The first acquisition unit 601 is used to acquire the initial action video of the opening and closing of the circuit breaker, wherein the opening and closing of the circuit breaker is used to characterize the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of the operating mechanism in the circuit breaker controlling the circuit breaker to perform the switching action, and the operating mechanism is a component in the circuit breaker. The second acquisition unit 602 is used to perform image enhancement processing on multiple initial images in the initial action video respectively to obtain multiple target images, and generate a target action video of the opening and closing of the circuit breaker from the multiple target images, wherein the multiple target images correspond one-to-one to the multiple initial images. The generation unit 603 is used to perform target tracking processing on the target action video to generate the running trajectory of the circuit breaker in the target action video. The determination unit 604 is used to determine the target speed of the opening and closing of the circuit breaker based on the running trajectory.

[0201] According to an embodiment of the present invention, a computer-readable storage medium is also provided, which includes a stored program, wherein when the program is executed by a processor, the device where the readable storage medium is located is controlled to execute the method for determining the speed of circuit breaker opening and closing in the embodiment.

[0202] According to an embodiment of the present invention, a processor is further provided, the processor being used to run a program, wherein the method for determining the speed of opening and closing of a circuit breaker in the embodiment is executed when the program is run.

[0203] According to an embodiment of the present invention, a computer program product is also provided, including a computer program, and when the computer program is executed by a processor, the method for determining the speed of opening and closing of a circuit breaker in an embodiment of the present invention is implemented.

[0204] According to an embodiment of the present invention, an electronic device is also provided, the electronic device comprising a processor and a memory for storing instructions executable by the processor, wherein the processor is configured to execute instructions to implement the method for determining the speed of opening and closing of a circuit breaker in an embodiment of the present invention.

[0205] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0206] In the several embodiments provided by the present invention, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic, for example, the division of units can be a logical function division, and there can be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0207] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0208] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0209] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0210] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for determining the speed of opening and closing of a circuit breaker, characterized in that: include: Acquire an initial action video of the circuit breaker opening and closing, wherein the circuit breaker opening and closing is used to characterize the on / off state of the circuit in which the circuit breaker is located, and the initial action video is used to characterize the process of the operating mechanism in the circuit breaker controlling the circuit breaker to perform the switching action, and the operating mechanism is a component in the circuit breaker; Performing image enhancement processing on a plurality of initial images in the initial action video respectively to obtain a plurality of target images, and generating a target action video of the circuit breaker opening and closing with the plurality of target images, wherein the plurality of target images correspond one to one to the plurality of initial images; Performing target tracking processing on the target action video to generate a running track of the circuit breaker in the target action video; Based on the operation trajectory, a target speed for opening and closing the circuit breaker is determined.

2. The method according to claim 1, characterized in that Get the initial action video of the circuit breaker opening and closing, including: An initial action video of the circuit breaker opening and closing taken by a target UAV is received, wherein the anti-electromagnetic interference capability of the receiver in the target UAV is higher than the anti-electromagnetic interference capability of the receiver in other UAVs.

3. The method according to claim 1, characterized in that Performing image enhancement processing on a plurality of initial images in the initial action video respectively to obtain a plurality of target images, and generating a target action video of the circuit breaker opening and closing from the plurality of target images, comprising: Performing frame interpolation processing on the multiple initial images in the initial action video respectively to obtain multiple first images, and generating a first action video of the circuit breaker opening and closing from the multiple first images, wherein the playback smoothness of the first action video is higher than the playback smoothness of the initial action video, and the multiple first images correspond to the multiple initial images one by one; performing dedistortion processing on the plurality of first images in the first action video respectively to obtain a plurality of second images, and generating a second action video of the circuit breaker opening and closing with the plurality of second images, wherein the distortion degree of the second images is lower than the distortion degree of the first images, and the plurality of second images correspond one-to-one to the plurality of first images; The image enhancement process is performed on the plurality of second images in the second action video respectively to obtain the target action video.

4. The method according to claim 3, characterized in that Performing the image enhancement process on the plurality of second images in the second action video respectively to obtain the target action video includes: Based on the detection requirement information of the circuit in which the circuit breaker is located, respectively determine a plurality of weight values ​​corresponding to a plurality of the second images; Based on the plurality of weight values, grayscale processing is performed on the plurality of second images respectively to obtain a plurality of third images, wherein the plurality of third images correspond one-to-one to the plurality of second images; The image enhancement process is performed on the plurality of third images respectively to obtain the target action video.

5. The method according to claim 4, characterized in that Performing the image enhancement processing on the plurality of third images respectively to obtain the target action video comprises: Performing histogram equalization processing on the plurality of third images respectively to obtain a plurality of fourth images, wherein the contrast of the fourth images is higher than the contrast of the third images, and the plurality of fourth images correspond to the plurality of third images one by one; Median filtering is performed on the multiple fourth images respectively to obtain multiple target images, wherein the overall image quality of the target images is higher than the overall image quality of the fourth images, and the multiple target images correspond to the multiple fourth images one by one.

6. The method according to claim 1, characterized in that Performing target tracking processing on the target action video to generate a running track of the circuit breaker in the target action video includes: Performing corner point detection on a plurality of the target images in the target action video to obtain target corner points of the plurality of the target images; The target corner point is input into a target tracking processing model for analysis to generate an operation trajectory of the circuit breaker, wherein the target tracking processing model is used to be established through target constraints, and the target constraints are used to maintain the continuity of the target action video.

7. The method according to claim 6, characterized in that The target constraint condition includes: a first constraint condition and a second constraint condition, wherein the first constraint condition is used to constrain the image brightness between adjacent frames in multiple target images respectively, and the second constraint condition is used to constrain the image conversion time between the adjacent frames.

8. A device for determining the speed of opening and closing of a circuit breaker, characterized in that: include: A first acquisition unit is used to acquire an initial action video of the circuit breaker opening and closing, wherein the circuit breaker opening and closing is used to represent the on-off state of the circuit in which the circuit breaker is located, and the initial action video is used to represent the process of the operating mechanism in the circuit breaker controlling the circuit breaker to perform the switching action, and the operating mechanism is a component in the circuit breaker; A second acquisition unit is used to perform image enhancement processing on the multiple initial images in the initial action video to obtain multiple target images, and generate a target action video of the circuit breaker opening and closing from the multiple target images, wherein the multiple target images correspond to the multiple initial images one by one; A generating unit, configured to perform target tracking processing on the target action video to generate a running track of the circuit breaker in the target action video; A determination unit is used to determine a target speed for opening and closing the circuit breaker based on the operation trajectory.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the device where the storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method according to any one of claims 1 to 7.

Images