A UAV, method and power inspection system for power inspection

Abstract

The present invention belongs to the technical field of electric power inspection drones, and provides a drone, method and electric power inspection system for electric power inspection. The drone includes a power system, a control system and a communication system; the control system communicates with a remote monitoring terminal through the communication system; the control system includes an obstacle avoidance module, an image acquisition module, a front-end image processing module and a main control module; the obstacle avoidance module is used to sense obstacles in front of the drone based on high-frequency millimeter wave signals and transmit them to the main control module, and the main control module forms an obstacle avoidance inspection path; the image acquisition module is mounted on the drone, and is used to collect electric power inspection images in the obstacle avoidance inspection path and transmit them to the front-end image processing module; the front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the center position of the image, and automatically adjust the exposure according to the histogram of the corrected image frame.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power inspection UAVs, and in particular relates to an UAV, method and electric power inspection system applied to electric power inspection. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] As an emerging tool for power inspection, drones are playing an increasingly important role in the intelligentization of power grids. Compared with traditional manual inspections, drone inspections have the following advantages: drone inspections have strong adaptability to the working environment and can overcome the harsh conditions of the working environment. In some complex environments, such as natural disasters such as ice and snow, earthquakes, and landslides, drones make up for the inconvenience of inspection personnel in transportation and can quickly operate in the on-site environment. Drone inspections replace manual inspections, reduce labor intensity and operating costs, make up for the problems of fatigue and visual blind spots in manual inspections, improve the accuracy and speed of inspections, and thus improve the efficiency of power inspections.

[0004] However, the inventors found that the current drones used for power inspections have the following defects:

[0005] (1) Although drones used for power inspections are equipped with conventional obstacle avoidance methods such as ultrasound and vision, their ability to perceive small objects is limited. When encountering objects such as wires and short branches, ultrasound and vision are often difficult to detect, causing drones to exceed the safe distance from the wires, resulting in consequences such as collisions with the wires, causing damage to personal and equipment safety.

[0006] (2) The shooting points during autonomous inspections by drones are completely dependent on the previous route planning. There are deviations in the shooting points of some power components, resulting in the power components to be photographed being incomplete in the image or being located at the edge of the image. Due to the uncertain inspection time and environmental weather, the captured images may be overexposed or underexposed, which will have a direct impact on the subsequent image analysis results. Summary of the invention

[0007] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a drone, method and power inspection system for power inspection, which can improve the drone's perception of small obstacles and significantly improve image quality, laying a good foundation for subsequent image analysis.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] A first aspect of the present invention provides a drone for power inspection.

[0010] A drone used for power inspection, comprising a power system, a control system and a communication system; the control system communicates with a remote monitoring terminal through the communication system; the control system comprises an obstacle avoidance module, an image acquisition module, a front-end image processing module and a main control module;

[0011] The obstacle avoidance module is used to sense obstacles in front of the drone based on high-frequency millimeter wave signals and transmit them to the main control module, which forms an obstacle avoidance inspection path;

[0012] The image acquisition module is mounted on the drone and is used to collect power inspection images in the obstacle avoidance inspection path and transmit them to the front-end image processing module;

[0013] The front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and to automatically adjust the exposure according to the histogram of the image frame after correction.

[0014] As an implementation manner, the obstacle avoidance module includes:

[0015] The millimeter-wave radar antenna is located at the front of the drone;

[0016] The radio frequency transmitter is used to modulate the signal emitted by the radar signal source into an intermediate frequency signal, and then up-convert it into a high-frequency millimeter wave signal, which is then amplified and radiated out through the millimeter wave radar antenna;

[0017] The receiver is used to receive the echo through the millimeter wave radar antenna, amplify it and mix it with the electromagnetic wave generated by the local oscillator, and obtain the intermediate frequency signal after amplitude amplification; one path of the intermediate frequency signal is envelope detected to determine whether there is an obstacle in the current direction, and the other path is used to sense the obstacle and automatically generate warning information.

[0018] As an implementation method, the millimeter wave radar antenna is in the form of a patch antenna array.

[0019] As an implementation method, the feeding amplitude and phase of each patch antenna determines the steering of the beam.

[0020] As an implementation mode, the millimeter wave radar antenna is also connected to a transceiver conversion switch, and the transceiver conversion switch is used to control the millimeter wave radar antenna to transmit or receive electromagnetic waves.

[0021] As an implementation mode, the millimeter wave radar antenna includes a feed antenna and an artificial metamaterial planar lens. The feed antenna uses an antenna operating in the 77 GHz frequency band, and the artificial metamaterial planar lens is used to converge and diverge electromagnetic waves.

[0022] As an implementation method, the front-end image processing module includes:

[0023] The target recognition submodule is used to determine the type of power component to be photographed according to the component name of the drone shooting point, and call the pre-trained target recognition model to perform target recognition on the image frames in the video stream;

[0024] The target determination submodule is used to extract all recognition results of the types of power components to be photographed, calculate the area size of each target, take the target with the largest area as the final target and confirm its coordinate information;

[0025] The image correction submodule is used to calculate the pixel-level offset of the final target position from the center of the image, so as to control the gimbal to adjust the angle so that the target is located in the center of the lens to achieve image correction;

[0026] The exposure adjustment submodule is used to determine the exposure state of the current image frame according to the histogram of the current image frame after correction, and then adjust the camera exposure to make the image frame brightness within a normal range.

[0027] As an implementation mode, in the image correction submodule, based on the pixel-level offset of the final target position from the image center position, according to the relationship between the image coordinate system, the camera coordinate system and the world coordinate system, the adjustment parameters of the gimbal are obtained.

[0028] As an implementation manner, in the exposure adjustment submodule, the exposure status includes normal exposure, overexposure and underexposure.

[0029] As an implementation manner, in the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deskew is distributed on the left side, the exposure state of the current image frame is underexposure.

[0030] As an implementation manner, in the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deskew is distributed on the right side, the exposure state of the current image frame is overexposure.

[0031] As an implementation manner, in the exposure adjustment submodule, if the histogram distribution of the current image frame after deflection correction is balanced, the exposure state of the current image frame is normal exposure.

[0032] As an implementation method, the communication system includes a 4G / 5G module.

[0033] As an implementation mode, the communication system further has a first security chip built in, and the first security chip establishes a trusted connection communication with the drone management and control platform through a secure access gateway.

[0034] As an implementation method, the drone management and control platform also establishes trusted connection communications with the drone ground control station and the remote control terminal respectively, and the drone ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

[0035] As an implementation method, the drone used for power inspection also establishes trusted connection communications with the drone ground control station and the remote control terminal respectively, and the drone ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

[0036] A second aspect of the present invention provides a working method of a drone used for power inspection.

[0037] A working method of a drone applied to power inspection as described above comprises:

[0038] The high-frequency millimeter wave signal of the obstacle avoidance module is used to sense obstacles in front of the drone and transmit them to the main control module, which then forms an obstacle avoidance inspection path.

[0039] The drone conducts power inspection according to the obstacle avoidance inspection path, uses the image acquisition module to collect power inspection images and transmits them to the front-end image processing module;

[0040] The front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and the exposure is automatically adjusted according to the histogram of the image frame after correction.

[0041] A third aspect of the present invention provides a power inspection system.

[0042] A power inspection system includes the drone used for power inspection as described above.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] (1) An innovative millimeter-wave radar obstacle avoidance technology was proposed, and an unmanned aerial vehicle (UAV) for power inspection was developed. The UAV uses a millimeter-wave radar to transmit the collected echo signal to the receiver and convert it into an intermediate frequency signal. The intermediate frequency signal processed by the receiver is divided into two paths, one for sensing the obstacle type and the other for determining the approximate location of the obstacle. This solves the problem of the limited perception of small objects by traditional unmanned power inspection vehicles, provides operators with more reaction time, realizes accurate perception of small obstacles, and improves the safety of the UAV.

[0045] (2) An innovative UAV front-end image correction and automatic exposure technology was proposed to solve the problems of incomplete shooting, overexposure, and underexposure of inspection image components. The technology controls the gimbal to adjust the angle so that the target is located in the center of the lens to achieve image correction based on the pixel-level offset of the final target position from the center of the image. The technology determines the exposure status of the current image frame based on the histogram of the current corrected image frame, and then adjusts the camera exposure to keep the image frame brightness within the normal range, thereby improving the quality of autonomous inspection images of UAVs and laying a foundation for subsequent image analysis and defect recognition.

[0046] (3) An innovative secure and trusted encryption communication technology for power inspection UAVs was proposed, and a secure and trusted encryption communication system for power inspection UAVs was developed. This solved the problem of no secure encryption measures for data between UAVs and additional security modules and the problem that the UAV control system could not complete the task independently. Based on the built-in security chips of the power inspection UAVs, ground control terminals and remote control terminals and the secure access gateway on the backend server, a trusted connection was established between the power inspection UAVs, ground control terminals and remote control terminals and the UAV control platform, as well as a trusted connection between the power inspection UAVs and the ground control terminals and remote control terminals. This achieved trusted access between each terminal connection platform and each terminal, encryption of transmitted information, and centralized management of access rules by the UAV control platform, thereby improving the security of UAV data.

[0047] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding 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 improper limitations on the present invention.

[0049] Figure 1 This is a schematic diagram of the structure of a drone used for power inspection according to an embodiment of the present invention;

[0050] FIG2( a ) is a front view of a millimeter wave radar module according to an embodiment of the present invention;

[0051] FIG2( b ) is a cross-sectional view of a millimeter wave radar module according to an embodiment of the present invention;

[0052] FIG3( a ) is a front view of a feed antenna according to an embodiment of the present invention;

[0053] FIG3( b ) is a rear view of the feed antenna according to an embodiment of the present invention;

[0054] FIG4( a) is a top view of a planar lens according to an embodiment of the present invention;

[0055] FIG4( b) is a side view of a planar lens according to an embodiment of the present invention;

[0056] Figure 5 4 is a structural diagram of a millimeter wave radar module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0058] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0059] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0060] Reference Figure 1 This embodiment provides a UAV for power inspection, which includes a power system, a control system and a communication system; the control system communicates with a remote monitoring terminal through the communication system; the control system includes an obstacle avoidance module, an image acquisition module, a front-end image processing module and a main control module.

[0061] In addition, it should be noted that in addition to the above modules, the control system also includes an inertial measurement unit (IMU), a gimbal control module, various attitude and position sensor modules, etc.

[0062] In the specific implementation process, the obstacle avoidance module is used to sense obstacles in front of the drone based on high-frequency millimeter wave signals and transmit them to the main control module, which forms an obstacle avoidance inspection path.

[0063] Specifically, Figure 5 As shown, the obstacle avoidance module of this embodiment includes:

[0064] The millimeter-wave radar antenna is located at the front of the drone;

[0065] The radio frequency transmitter is used to modulate the signal emitted by the radar signal source into an intermediate frequency signal, and then up-convert it into a high-frequency millimeter wave signal, which is then amplified and radiated out through the millimeter wave radar antenna;

[0066] The receiver is used to receive the echo through the millimeter wave radar antenna, amplify it and mix it with the electromagnetic wave generated by the local oscillator, and obtain the intermediate frequency signal after amplitude amplification; one path of the intermediate frequency signal is envelope detected to determine whether there is an obstacle in the current direction, and the other path is used to sense the obstacle and automatically generate warning information.

[0067] In a specific implementation process, the millimeter wave radar antenna is in the form of a patch antenna array, and the feeding amplitude and phase of each patch antenna determine the steering of the beam.

[0068] As shown in FIG. 2( a ) and FIG. 2( b ), the millimeter-wave radar antenna includes a feed antenna 1 and an artificial metamaterial planar lens 2 . The feed antenna 1 uses an antenna operating in the 77 GHz frequency band, and the artificial metamaterial planar lens 2 is used to converge and diverge electromagnetic waves.

[0069] The feed antenna 1 uses an antenna array 4 (such as a 3*3 patch) operating in the 77 GHz frequency band, as shown in Figure 3(a) and Figure 3(b). In Figure 3(b), 5 is a feeding point.

[0070] The advantage of using a patch antenna array is that the patch antenna array has the characteristics of controllable beam and easy integration. Based on the antenna array synthesis theory, by adjusting the feeding amplitude and phase of each patch antenna, the beam can be steered to achieve the effect of phased scanning, thus mentioning the mechanical rotating gimbal used in traditional radars.

[0071] The feed antenna can theoretically use any form of 77GHz band antenna. For example, a general lens antenna often uses a horn antenna as a feed antenna because the horn antenna has the characteristics of strong directivity, high gain, simple structure and easy design. However, the horn antenna is a three-dimensional structure that is not easy to integrate and needs to be aligned with the center of the lens. For UAV platforms that are prone to vibration, the requirements for the mechanical fixing structure are too high, so it is not used. The patch array antenna used in this embodiment is a two-dimensional planar structure with a negligible thin thickness, a large main lobe beam width, less affected by vibration, and a large fault tolerance rate. The plane lens has certain requirements for the incident angle of the feed antenna. If the incident angle offset of the feed antenna is too large, the refraction effect of the lens will deteriorate. However, if the feed antenna has no incident angle deviation, a mechanical rotation structure is required to assist the antenna in scanning. The structure is bulky and the effect is not good. Therefore, a 3*3 array is used to form a small phased array. The scanning angle of the feed antenna can be slightly offset by changing the amplitude phase of the array element (the more array elements there are, the larger the scanning angle that can be achieved without much reduction in gain), and the scanning angle can be limited to the acceptable incident angle offset range of the plane lens. The 3*3 array proposed in this embodiment can produce a scanning angle of 60 degrees, which meets the incident angle requirement of the plane lens and retains a certain margin.

[0072] In this embodiment, as shown in FIG4(a) and FIG4(b), the artificial metamaterial plane lens 2 has advantages such as small size and light weight compared to the traditional Luneburg lens and gradient refractive index lens, which makes it easier to integrate into application scenarios with limited space. According to Maxwell's equations, the propagation speed of electromagnetic waves in the medium is Where ε and μ are the effective dielectric constant and effective magnetic permeability of the medium, ε0 and μ0 are the vacuum dielectric constant and vacuum magnetic permeability, ε r and μ r is the relative dielectric constant and relative magnetic permeability. It can be seen that the propagation speed is only related to the relative dielectric constant and relative magnetic permeability of the material. If these two properties can be changed, the direction of electromagnetic wave propagation can be controlled. Artificial electromagnetic metamaterials are composed of a "medium + conductor" structure. The relative dielectric constant and relative magnetic permeability of the metamaterial structure can be adjusted by changing the shape of the conductor. According to the definition of the refractive index of electromagnetic waves Where c is the speed of light in a vacuum. It can be seen that the relative dielectric constant and relative magnetic permeability can further affect the refractive index, thereby realizing the convergence and divergence of electromagnetic waves.

[0073] In the specific implementation process, the feed antenna 1 of the millimeter wave radar adopts a 3*3 patch antenna array, the patch antenna array element size is 2.1mm*1.8mm, the array element spacing is 0.6mm, the size of the entire feed antenna is 13mm*10mm, the focal length of the plane lens is selected to be 30mm, the lens is composed of 2mm*2mm metamaterial units of equal thickness, the size of the entire lens is 50mm*50mm, and there are 25*25 "medium + conductor" structural units. Each "medium + conductor" structural unit adopts a 3-layer medium + conductor structure to achieve a sufficient refractive index.

[0074] The number of layers of the unit is related to the transmission coefficient and refraction angle range of the lens. The more layers there are, the smaller the transmission coefficient of the lens, the less energy the electromagnetic wave can pass through the lens, the lower the efficiency, but the larger the refraction angle range. Therefore, a compromise needs to be considered during design. The three layers of "medium + conductor" in this embodiment meet the refractive index requirements, so a three-layer structure is adopted.

[0075] The feed antenna 1 and the artificial metamaterial planar lens 2 of this embodiment both form a conformal structure with the drone body, and their modules are integrated on a circuit board inside the drone shell 3. Their communication serial port is connected to the reserved I / O port of the flight control module, supporting the setting of parameters such as the scanning path and period of the millimeter-wave radar through the flight control module.

[0076] The RF transmitter, receiver and transceiver conversion switch control the millimeter-wave radar antenna to transmit or receive electromagnetic waves. The RF transmitter can modulate the signal emitted by the radar signal source into an intermediate frequency signal, and then up-convert it to form a high-frequency millimeter-wave signal, which is amplified by the RF power amplifier and radiated through the antenna; the transceiver conversion switch uses a millimeter-wave duplexer with a substrate integrated waveguide structure, which has low insertion loss, high transceiver isolation and is easy to integrate; the receiver receives the echo through the radar antenna, passes through the receiver protector, low-noise amplifier, and is mixed with the electromagnetic wave generated by the local oscillator, and then passes through the intermediate frequency amplifier to obtain an intermediate frequency signal with an amplitude amplification.

[0077] The transmitted radar signal is a frequency modulated continuous wave, which can be expressed as:

[0078]

[0079] Where A is the signal amplitude, T is the transmission time width, f0 is the base frequency 77GHz, and the transmission signal sweep bandwidth is recorded as B. is the frequency modulation slope, is the initial phase.

[0080] Then the phase of the transmitted signal can be written as

[0081] Assuming the distance between the target and the radar is R and the propagation speed of electromagnetic waves is c, the received signal can be written as

[0082] Where K is the signal attenuation and τ is the received signal delay, the phase of the echo signal can be expressed as Then the phase of the difference frequency signal between it and the transmitted signal can be expressed as p t (t)-p r (t)=2πf0τ+2πuτt-πuτ 2 , where the first and third terms are constants, then the frequency of the intermediate frequency signal can be obtained The distance can be expressed as

[0083] The intermediate frequency signal is input to the envelope detector. The envelope detector can be used to detect the envelope change of the echo. When there is an obstacle in front, the envelope of the echo will fluctuate with a peak. The direction of the obstacle is determined by the direction of the echo. The traditional radar azimuth scanning usually uses a mechanical rotating pan-tilt to drive the antenna to periodically scan a certain area along a fixed path. The millimeter wave radar feed antenna of the present invention uses a 3*3 patch antenna array. According to the antenna array synthesis principle, the radiation beam can be adjusted by controlling the feed amplitude and phase of each patch array element. The beam is irradiated on the plane lens for further refraction, and a narrow beam with stronger directivity can be obtained. The narrow beam is controlled by the feed amplitude and phase of the feed antenna, thereby realizing electronic scanning.

[0084] The distance between the obstacle and the drone can be estimated by the frequency of the intermediate frequency signal. These azimuth information and envelope information together constitute a video signal of the radar. Specifically, the radar obtains the echo information within its observable range through electronic scanning. Assuming that there is an obstacle at a certain point, its echo will have a peak after envelope detection. After the peak is captured, the corresponding radar beam direction where the peak appears is also marked (obtained by calculating the amplitude and phase characteristics of the feed antenna array unit). These two pieces of information are combined and displayed on the operator's display screen, which can clearly show the approximate direction of the obstacle and give the operator time to react without waiting for the algorithm module to process the intermediate frequency signal before displaying it. At this time, the operator cannot judge the specific shape of the obstacle in that direction, but can only know that there is an obstacle in that direction.

[0085] The other intermediate frequency signal is input into the algorithm module for analysis. Since the sensed objects are often small objects such as branches and wires, the traditional CFAR algorithm filters the echo signal to form a radar point cloud signal processing method is not advisable. Therefore, it is decided to retain dense data and use deep learning to extract the features of wires and branches. A neural network is used to analyze the features of obstacles and gradually form a sample library of echo features of obstacles such as wires and branches, thus achieving reliable obstacle recognition and automatic obstacle avoidance.

[0086] For millimeter-wave radar, a wire can be equivalent to a line composed of countless points. Each point on the wire does not exist independently. Assuming that the radar resolution divides the detection area into countless boxes in space, the wire echo must be continuous in one direction and discontinuous in its normal direction. (Assuming that the wire is horizontal, the wire echo is continuous in the horizontal direction and discontinuous in the vertical direction. This is the characteristic of the wire echo). The wire echo information is put into the training set for training, the machine learning model parameters are corrected, and the wire features are extracted. After large-scale sample training, the model can judge whether the input echo information is a wire, and the same applies to obstacles such as leaves.

[0087] When an echo with an amplitude exceeding the set threshold appears in a certain direction, the algorithm chip first extracts the echo features, marks the direction, and continues to analyze the locations around the direction in the next few frames of radar signals. When the echo is detected to disappear or deviate, the tracking of this direction is abandoned. When the echo position is detected to be close to the safe distance, the algorithm chip automatically generates an alarm message to prompt the operator. When the echo position is detected to be about to exceed the safe distance, the algorithm chip sends an obstacle avoidance command to the UAV flight control module to guide the UAV to execute the set obstacle avoidance action, thereby ensuring the safe flight of the UAV.

[0088] Among them, the safety distance is a parameter that can be set on the drone control platform. For example, if it is 5m, an obstacle avoidance command will be sent when the distance between the detected echo and the drone is close to 5m.

[0089] In the specific implementation process, the image acquisition module is mounted on the drone and is used to collect power inspection images in the obstacle avoidance inspection path and transmit them to the front-end image processing module.

[0090] Specifically, the image acquisition module can be implemented by a camera, such as an infrared camera, a visible light camera, etc.

[0091] In a specific implementation process, the front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and automatically adjust the exposure according to the histogram of the corrected image frame.

[0092] Wherein, the front-end image processing module includes:

[0093] The target recognition submodule is used to determine the type of power component to be photographed according to the component name of the drone shooting point, and call the pre-trained target recognition model to perform target recognition on the image frames in the video stream;

[0094] The target determination submodule is used to extract all recognition results of the types of power components to be photographed, calculate the area size of each target, take the target with the largest area as the final target and confirm its coordinate information;

[0095] The image correction submodule is used to calculate the pixel-level offset of the final target position from the center of the image, so as to control the gimbal to adjust the angle so that the target is located in the center of the lens to achieve image correction;

[0096] The exposure adjustment submodule is used to determine the exposure state of the current image frame according to the histogram of the current image frame after correction, and then adjust the camera exposure to make the image frame brightness within a normal range.

[0097] Specifically, in the image correction submodule, based on the pixel-level offset of the final target position from the image center position, according to the relationship between the image coordinate system, the camera coordinate system and the world coordinate system, the adjustment parameters of the gimbal are obtained.

[0098] In the exposure adjustment submodule, the exposure status includes normal exposure, overexposure and underexposure.

[0099] In the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deflection correction is distributed on the left side, the exposure state of the current image frame is underexposure.

[0100] In the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deflection correction is distributed on the right side, the exposure state of the current image frame is overexposure.

[0101] In the exposure adjustment submodule, if the histogram distribution of the current image frame after deflection correction is balanced, the exposure state of the current image frame is normal exposure.

[0102] In some other embodiments, the communication system includes a 4G / 5G module.

[0103] Among them, the 4G / 5G module is used to transmit key data information of the drone through the 4G / 5G link.

[0104] For example, the current channel between the drone control terminal and the drone body is radio, and a certain transmission power cannot exceed the limit distance. When long-distance communication exceeds the limit distance, the control command is switched to the 4G / 5G communication network. When the control terminal and the drone meet the radio communication conditions, the controller authority is transferred back to radio communication. This solution can greatly expand the flight control distance of the drone, break through the existing limit, and meet the long-distance communication needs in complex environments such as mountain / building obstruction during power inspection.

[0105] During the flight, the drone determines whether it is close to the control command limit distance by the received control end radio signal strength (control command resolution sensitivity threshold). When it is close to the limit distance, the drone sends a request to the control end to switch the data transmission link. After the control end confirms, the 4G / 5G module (the drone is in a dormant state when flying normally) starts, registers the 5G network (registration process), and measures the 5G network signal strength. When the measurement meets the drone's stable access threshold (measure whether the main service 5G area strength meets a certain threshold in a certain period of time), the control end is again asked by radio signal whether to switch the control command channel to the 5G network. Once the control end instructs the drone to switch the control command path to the 5G network (for drone control commands, 5G ultra-reliable low-latency communication URLLC must be enabled, and the corresponding slice data must be configured to ensure high-reliability and low-latency control of the drone), the control end and the drone 5G module access the operator's 5G network and complete the login to the drone control platform. The control end and the drone turn off the radio communication module, and the subsequent control command path is completely transmitted by the 5G network to the drone control platform for processing and carrying.

[0106] After the control command channel is switched to the 5G network, the control end can know the position of the drone in real time through the 5G network. The drone monitors the distance from the control end in real time through its own positioning module. Once the distance is less than a certain threshold (such as the current direct control limit of 7km) within a certain period of time, the drone reports the request to turn on the radio module to the control end through the 5G network. The control end sends a confirmation to allow the radio module to be turned on through the 5G network, and autonomously turns on the radio signal module of the control end to send a handshake test signal. After receiving the confirmation, the drone autonomously turns on the radio module and starts scanning the handshake test signal of the control end. If the handshake test signal maintains stable strength and quality within a certain period of time (meeting the radio signal analysis sensitivity), the drone asks the control end whether to switch back to the direct connection mode (radio communication) through the 5G network. The control end sends a confirmation to switch back through the 5G network, and at the same time receives the drone's switch back signal at the radio signal end. Once the drone receives the switch back notification in the radio signal, the control end turns off the 5G module (deregistration process). After the drone receives confirmation of switching back to radio communication on the 5G network, it transmits the switch back notification through the radio module and turns off the 5G module (deregistration process). The control end and the drone enter the radio direct control stage again.

[0107] In other embodiments, the communication system further has a first security chip built in, and the first security chip establishes a trusted connection communication with the drone management and control platform through a secure access gateway.

[0108] Among them, the drone management and control platform also establishes trusted connection communications with the drone ground control station and the remote control terminal respectively, and the drone ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

[0109] The drone used for power inspection also establishes trusted connection communications with the drone ground control station and the remote control terminal respectively, and the drone ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

[0110] The first security chip, the second security chip and the third security chip of this embodiment provide support for the operation of national secret SM1 / 2 / 3 / 4 and other algorithms, and securely store confidential information such as keys and digital certificates. The first security chip, the second security chip and the third security chip also support high-speed communication interfaces, with low data processing latency and high throughput, reducing the negative impact of adding national secret algorithms on the original power drone business process. It has comprehensive protection technology and high security, which can prevent various types of intrusion and theft attacks and ensure the security of the entire upper-layer service.

[0111] In this embodiment, the power system includes an electronic speed regulator, a propeller, a motor and a power supply. The specific structure thereof can be selected by those skilled in the art according to actual conditions and will not be described in detail here.

[0112] In one or more embodiments, a working method of a drone applied to power inspection as described above is provided, which includes:

[0113] Step 1: Use the high-frequency millimeter wave signal of the obstacle avoidance module to sense the obstacles in front of the drone and transmit it to the main control module, which forms an obstacle avoidance inspection path;

[0114] Step 2: The drone performs power inspection according to the obstacle avoidance inspection path, uses the image acquisition module to collect power inspection images and transmits them to the front-end image processing module;

[0115] Step 3: Use the front-end image processing module to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and automatically adjust the exposure according to the histogram of the corrected image frame.

[0116] In one or more embodiments, a power inspection system is also provided, which includes the drone used for power inspection as described above.

[0117] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A drone used for power inspection, comprising a power system, a control system and a communication system; the control system communicates with a remote monitoring terminal through the communication system; It is characterized in that The control system includes an obstacle avoidance module, an image acquisition module, a front-end image processing module and a main control module; The obstacle avoidance module is used to sense obstacles in front of the drone based on high-frequency millimeter wave signals and transmit them to the main control module, which forms an obstacle avoidance inspection path; The obstacle avoidance module comprises: The millimeter-wave radar antenna is located at the front of the drone; The radio frequency transmitter is used to modulate the signal emitted by the radar signal source into an intermediate frequency signal, and then up-convert it into a high-frequency millimeter wave signal, which is then amplified and radiated out through the millimeter wave radar antenna; The receiver is used to receive the echo through the millimeter wave radar antenna, amplify it and mix it with the electromagnetic wave generated by the local oscillator, and obtain the intermediate frequency signal after the amplitude is amplified; one path of the intermediate frequency signal is envelope-detected to determine whether there is an obstacle in front of it. When there is an obstacle in front of it, the envelope of the echo will have a peak fluctuation. The other path is used to sense the obstacle and automatically generate warning information; The millimeter wave radar antenna is in the form of a patch antenna array, and the feeding amplitude and phase of each patch antenna determine the steering of the beam; based on the antenna array synthesis theory, the beam steering is achieved by adjusting the feeding amplitude and phase of each patch antenna; The millimeter-wave radar feed antenna adopts a 3*3 patch antenna array. According to the antenna array synthesis principle, the radiation beam is adjusted by controlling the feed amplitude and phase of each patch array element. The beam is irradiated on the plane lens and further refracted to obtain a narrow beam with stronger directivity. The narrow beam is controlled by the feed amplitude and phase of the feed antenna to achieve electronic scanning. The distance between the obstacle and the UAV is estimated by the frequency of the intermediate frequency signal, and the azimuth information and envelope information together constitute a video signal of the radar. Specifically, the radar obtains the echo information within the observation range through electronic scanning. Assuming that there is an obstacle at a certain point, its echo will have a peak after envelope detection. After the peak is captured, the corresponding radar beam direction where the peak appears is also marked, which is obtained by calculating the amplitude and phase characteristics of the feed antenna array unit. These two pieces of information are combined to display the approximate direction of the obstacle on the display screen, which clearly shows the approximate direction of the obstacle and gives the operator time to react. After the peak is captured, the corresponding radar beam direction where the peak appears is also marked, which is obtained by calculating the amplitude and phase characteristics of the feed antenna array unit; The other intermediate frequency signal is input into the algorithm module for analysis. The deep learning method is used to extract the characteristics of wires and branches. The neural network is used to analyze the characteristics of obstacles and gradually form a sample library of wire and branch obstacle echo characteristics, realizing reliable obstacle recognition and automatic obstacle avoidance functions. When an echo with an amplitude exceeding the set threshold appears in a certain direction, the algorithm chip first extracts the echo feature, marks the direction, and continuously analyzes the locations around the direction in the next few frames of radar signals. When the echo disappears or deviates, the tracking of this direction is abandoned. When the echo position is detected to be close to the safe distance, the algorithm chip automatically generates an alarm message to prompt the operator. When the echo position is detected to be about to exceed the safe distance, the algorithm chip sends an obstacle avoidance command to the UAV flight control module to guide the UAV to execute the set obstacle avoidance action, thereby ensuring the safe flight of the UAV. Among them, the safe distance is a parameter that can be set on the UAV control platform. The image acquisition module is mounted on the drone and is used to collect power inspection images in the obstacle avoidance inspection path and transmit them to the front-end image processing module; The front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and automatically adjust the exposure amount according to the histogram of the image frame after correction; wherein, based on the pixel-level offset between the position of the final target and the image center position, the adjustment parameters of the gimbal are obtained according to the relationship between the image coordinate system, the camera coordinate system and the world coordinate system; The front-end image processing module includes: The target recognition submodule is used to determine the type of power component to be photographed according to the component name of the drone shooting point, and call the pre-trained target recognition model to perform target recognition on the image frames in the video stream; The target determination submodule is used to extract all recognition results of the types of power components to be photographed, calculate the area size of each target, take the target with the largest area as the final target and confirm its coordinate information; The image correction submodule is used to calculate the pixel-level offset of the final target position from the center of the image, so as to control the gimbal to adjust the angle so that the target is located in the center of the lens to achieve image correction; An exposure adjustment submodule, which is used to determine the exposure state of the current image frame according to the histogram of the current image frame after correction, and then adjust the camera exposure so that the brightness of the image frame is within a normal range; wherein the exposure state includes normal exposure, overexposure and underexposure; Based on the built-in security chips of the power inspection UAV, ground control terminal and remote control terminal and the secure access gateway on the backend server, a trusted connection is established between the power inspection UAV, ground control terminal and remote control terminal and the UAV management and control platform, as well as a trusted connection between the power inspection UAV and the ground control terminal and remote control terminal, to achieve trusted access and encryption of transmission information between each terminal connection platform and each terminal, and centralized management of access rules by the UAV management and control platform; The UAV control platform also establishes trusted connection communications with the UAV ground control station and the remote control terminal respectively; The UAV used for power inspection also establishes trusted connection communications with the UAV ground control station and the remote control terminal respectively.

2. The drone used for power inspection as claimed in claim 1, It is characterized in that The millimeter wave radar antenna is also connected to a transceiver conversion switch, and the transceiver conversion switch is used to control the millimeter wave radar antenna to transmit or receive electromagnetic waves.

3. The drone used for power inspection as claimed in claim 1, It is characterized in that The millimeter wave radar antenna includes a feed antenna and an artificial metamaterial planar lens. The feed antenna uses an antenna working in the 77 GHz frequency band, and the artificial metamaterial planar lens is used to converge and diverge electromagnetic waves.

4. The drone used for power inspection as claimed in claim 1, It is characterized in that In the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deflection correction is distributed on the left side, the exposure state of the current image frame is underexposure.

5. The drone used for power inspection as claimed in claim 1, It is characterized in that In the exposure adjustment submodule, if more than 60% of the histogram of the current image frame after deflection correction is distributed on the right side, the exposure state of the current image frame is overexposure.

6. The drone used for power inspection as claimed in claim 1, It is characterized in that In the exposure adjustment submodule, if the histogram distribution of the current image frame after deflection correction is balanced, the exposure state of the current image frame is normal exposure.

7. The drone used for power inspection as claimed in claim 1, It is characterized in that The communication system includes a 4G / 5G module.

8. The drone used for power inspection as claimed in claim 1, It is characterized in that The communication system also has a first security chip built in, and the first security chip establishes a trusted connection communication with the drone management and control platform through a secure access gateway.

9. The drone used for power inspection as claimed in claim 8, It is characterized in that The UAV ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

10. The drone used for power inspection as claimed in claim 8, It is characterized in that The UAV ground station and the remote control terminal are respectively provided with a second security chip and a third security chip.

11. A working method of a drone used for power inspection according to any one of claims 1 to 10, It is characterized in that include: The high-frequency millimeter wave signal of the obstacle avoidance module is used to sense obstacles in front of the drone and transmit them to the main control module, which then forms an obstacle avoidance inspection path. The drone conducts power inspection according to the obstacle avoidance inspection path, uses the image acquisition module to collect power inspection images and transmits them to the front-end image processing module; The front-end image processing module is used to perform image correction according to the pixel-level offset between the target position in the image frame and the image center position, and the exposure is automatically adjusted according to the histogram of the image frame after correction.

12. A power inspection system, It is characterized in that It comprises a drone used for power inspection as described in any one of claims 1-10.

Images