Method, device and system for controlling the height of flight, and cleaning system

Abstract

The application discloses a flight height control method and device and a cleaning system, and belongs to the technical field of photovoltaic cleaning. Image data and point cloud data collected in real time during the flight of a UAV are acquired, and a target cleaning distance between the UAV and a target component is determined in real time according to the image data and the point cloud data. Since the image data and the point cloud data can complementarily represent the position information of the target component from two dimensions of visual semantics and spatial geometry, the target cleaning distance determined according to the image data and the point cloud data can accurately reflect the spatial position relationship between the UAV and the target component, so that the flight height of the UAV is dynamically adjusted according to the target cleaning distance, the UAV can maintain a safe and effective working distance from the surface of the target component in the cleaning operation, and thus precise adjustment of the flight height is realized.

Description

Technical Field

[0001] This application relates to the field of photovoltaic cleaning technology, specifically to flight altitude control methods, devices, and cleaning systems. Background Technology

[0002] Using drones equipped with cleaning devices to clean photovoltaic modules offers advantages such as high efficiency, good safety, wide coverage, and no mechanical damage to the modules. However, the effectiveness of photovoltaic module cleaning drones is directly related to the stability and accuracy of their flight altitude. For example, if the flight altitude is too high, the cleaning device cannot effectively contact the module surface, failing to achieve the desired cleaning effect. If the flight altitude is too low, the drone and cleaning device are prone to collisions with the photovoltaic modules, causing scratches and damage to the drone, among other safety hazards. Therefore, precise adjustment of flight altitude is one of the core technologies for achieving efficient and safe operation of photovoltaic module cleaning drones. Summary of the Invention

[0003] This application provides a flight altitude control method, apparatus, and cleaning system to achieve precise adjustment of the flight altitude of a photovoltaic cleaning drone.

[0004] Firstly, a flight altitude control method is provided for use with a drone for cleaning photovoltaic modules; the method includes: Acquire image and point cloud data collected during the drone's flight; Based on image data and point cloud data, determine the target cleaning distance between the drone and the target component; Adjust the drone's flight altitude according to the target cleaning distance.

[0005] Secondly, a flight altitude control device is also provided for use in drones cleaning photovoltaic modules; the device includes: The acquisition module is used to acquire image data and point cloud data collected during the flight of the UAV; The determination module is used to determine the target cleaning distance between the UAV and the target component based on the image data and the point cloud data; The adjustment module is used to adjust the drone's flight altitude based on the target cleaning distance.

[0006] Thirdly, a cleaning system for photovoltaic modules is provided, comprising: Drones used for cleaning photovoltaic modules; Sensors mounted on the underside of the drone are used to collect image data and point cloud data during the drone's flight. The computing power box is used to determine the target cleaning distance between the drone and the target components based on image data and point cloud data; The control module is used to adjust the drone's flight altitude based on the target cleaning distance.

[0007] This application provides a flight altitude control method, apparatus, and cleaning system. The flight altitude control method provided in this application acquires real-time image data and point cloud data collected during the flight of a drone, and determines the target cleaning distance between the drone and the target component in real time based on the image data and point cloud data. Since image data and point cloud data can complementaryly represent the positional information of the target component from both visual semantic and spatial geometric dimensions, the target cleaning distance determined based on the image data and point cloud data can accurately reflect the spatial positional relationship between the drone and the target component. Therefore, dynamically adjusting the drone's flight altitude based on the target cleaning distance enables the drone to maintain a safe and effective working distance from the target component surface during cleaning operations, thereby achieving precise adjustment of the flight altitude. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic diagram of the structure of a cleaning system provided in an embodiment of this application; Figure 2 This is a flowchart of a flight altitude control method provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure and location of a target component provided in an embodiment of this application; Figure 4 This is a top view of the target component provided in the embodiments of this application; Figure 5 This is a schematic diagram illustrating the target parameter annotation of a target component provided in an embodiment of this application; Figure 6 This is a schematic diagram of the fitting reference surface provided in an embodiment of this application; Figure 7 This is a schematic diagram of the spatial envelope provided in an embodiment of this application; Figure 8 This is a schematic diagram of the overall process of a flight altitude control method provided in an embodiment of this application; Figure 9 This is a schematic diagram of the flight altitude control device provided in an embodiment of this application. Detailed Implementation

[0010] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] In the description of this application, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0012] "A and / or B" includes the following three combinations: A only, B only, and a combination of A and B.

[0013] The use of "applies to" or "configured to" in this application implies open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more of the stated conditions or values ​​may in practice be based on additional conditions or values ​​beyond those stated.

[0014] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0015] The applicant's research revealed that during the operation and maintenance of photovoltaic power plants, dust, bird droppings, debris, and other contaminants easily accumulate on the surface of photovoltaic modules, leading to a decrease in module transmittance and significantly affecting photovoltaic power generation efficiency. Therefore, regular cleaning of photovoltaic modules is a crucial step in ensuring the power generation efficiency of the power plant. With the rapid development of drone technology, using drones equipped with cleaning devices to clean photovoltaic modules has certain advantages over related technologies that rely on manual cleaning or ground-based mechanical cleaning. This is especially true for cleaning modules in large-scale photovoltaic power plants (particularly complex scenarios such as rooftop photovoltaics and desert photovoltaics), where it offers advantages such as high operational efficiency, good safety, wide coverage, and no mechanical damage to the modules.

[0016] The effectiveness of photovoltaic module cleaning drones is directly related to the stability and accuracy of their flight altitude. For example, if the flight altitude is too high, the cleaning device cannot effectively contact the module surface, failing to achieve the desired cleaning effect. If the flight altitude is too low, the drone fuselage and cleaning device are prone to collisions with the photovoltaic modules, causing scratches on the modules and damage to the drone, among other safety hazards. Therefore, precise adjustment of flight altitude is one of the core technologies for achieving efficient and safe operation of photovoltaic module cleaning drones.

[0017] Currently, the flight altitude adjustment technology for photovoltaic module cleaning drones can be mainly divided into the following solutions.

[0018] For example, related technologies provide an adjustment scheme based on a preset flight path and fixed altitude. This scheme involves surveying the photovoltaic power station beforehand to obtain the approximate distribution and elevation information of the modules, preset the flight path and fixed altitude of the drone, and during operation, the drone flies at a constant speed according to the preset parameters without real-time altitude adjustments. Some of these schemes are paired with simple distance sensors (such as ultrasonic sensors) to make coarse altitude corrections only when abnormal distances are detected. This type of scheme relies on the preliminary survey data and is suitable for small photovoltaic power stations with neatly arranged modules and flat ground. However, it cannot handle complex scenarios, especially in actual photovoltaic power stations where modules may be installed at tilt or varying heights, and the ground is often uneven (such as the undulating terrain of desert photovoltaic power stations). The preset fixed altitude cannot adapt to the actual height changes of the modules, which can easily lead to incomplete cleaning or collisions with the modules.

[0019] For example, some technologies rely on ground-based altitude-fixed flight, detecting and adjusting the distance between the drone and the ground in real time. This approach only considers the height relative to the ground, not the relative height to the components. Its drawback is that the distance between the drone and the components may not be optimal, potentially leading to excessive distance that negatively impacts cleaning effectiveness.

[0020] For example, some technologies rely on photovoltaic panels to maintain a fixed altitude during flight, and only the height of the photovoltaic panels is detected during cleaning. This approach may cause the ground to be mistaken for the panels when encountering gaps between components, leading to a risk of collision.

[0021] For example, related technologies propose a flight control method for photovoltaic cleaning drones. This method first determines the terrain scene based on the spacing of the modules and the slope of the power station, and then uses different methods to adjust the altitude for different scenes. Moreover, the methods for detecting ground altitude are also different for different scenes, either using only radar detection or only image detection.

[0022] For example, related technologies propose a method that combines visual and laser positioning to determine the real-time pose of a drone, presets inspection coordinates and altitudes for the drone to define the inspection area, and adjusts the drone's flight altitude based on the determined wind speed. This method adjusts the altitude based on the preset parameters and wind speed.

[0023] For example, related technologies propose a drone flight altitude adjustment system that uses a spherical camera on the drone to image the ground scene, obtain and output ground scene images, and adjust the altitude based on the image information.

[0024] In view of this, embodiments of this application provide a flight altitude control method, apparatus, and cleaning system. This system acquires real-time image data and point cloud data collected during the flight of a drone, and determines the target cleaning distance between the drone and the target component in real time based on the image data and point cloud data. Since image data and point cloud data can complementaryly represent the positional information of the target component from both visual semantic and spatial geometric dimensions, the target cleaning distance determined based on the image data and point cloud data can accurately reflect the spatial positional relationship between the drone and the target component. Therefore, dynamically adjusting the drone's flight altitude based on the target cleaning distance enables the drone to maintain a safe and effective working distance from the target component surface during cleaning operations, thereby achieving precise adjustment of the flight altitude.

[0025] Figure 1 This is a schematic diagram of a cleaning system provided in an embodiment of this application. This application provides a cleaning system for photovoltaic modules, see below. Figure 1 The cleaning system 10 includes: a drone 11 for cleaning photovoltaic modules; a sensor 12 mounted below the drone 11 for collecting image data and point cloud data during the flight of the drone 11; a computing power box 13 for determining the target cleaning distance between the drone 11 and the target module based on the image data and point cloud data; and a control module 14 for adjusting the flight altitude of the drone 11 according to the target cleaning distance.

[0026] The drone 11 is used to carry a water tank for cleaning photovoltaic modules. The control module 14 can be a separate control chip or a control chip integrated into the drone 11, etc., depending on the actual situation, and is not specifically limited here. The sensor 12 is small in size and is arranged directly below the drone 11. The distance between the drone 11 and the target module is approximately equal to the distance between the sensor 12 and the target module. Therefore, in this embodiment, the target cleaning distance can be either the distance between the drone 11 and the target module or the distance between the sensor 12 and the target module, depending on the actual situation, and is not specifically limited here.

[0027] In some embodiments, continue reading Figure 1 Sensor 12 may include an image sensor 121, a point cloud data sensor 122, and an inertial measurement unit (IMU) sensor 123. The image sensor 121 is used to acquire image data during the flight of the drone 11, for example, capturing image data below the drone 11 to identify whether a target component has been detected. The point cloud data sensor 122 is used to acquire point cloud data during the flight of the drone 11, for example, capturing point cloud data below the drone 11 to acquire point cloud data of the target component. The IMU sensor 123 is used to calculate the direction of gravity (or geocentric direction) as a basis for identifying the ground. Exemplarily, the image sensor 121 may be a camera or module, and the point cloud data sensor 122 may be a lidar sensor.

[0028] As one implementation, the image sensor 121 and the point cloud data sensor 122 can also be a sensor that can simultaneously acquire image data and point cloud data, such as a binocular camera, a binocular structured light camera, a monocular structured light camera, and a time-of-flight camera (TOF), etc. The specific settings can be made according to the actual situation, and no specific limitations are made here.

[0029] Figure 2 This is a flowchart illustrating a flight altitude control method provided in an embodiment of this application. This application provides a flight altitude control method, see below. Figure 2 The method may include the following steps S100 to S300.

[0030] Step S100: Acquire image data and point cloud data collected during the flight of UAV 11; The image data may or may not include images of the target component. Similarly, the point cloud data may or may not include point cloud data of the target component. The point cloud data is 3D point cloud data.

[0031] Step S200: Determine the target cleaning distance between the UAV 11 and the target component based on the image data and point cloud data.

[0032] The target component is the photovoltaic module that the drone 11 is currently flying at to be determined for cleaning.

[0033] Figure 3 This is a schematic diagram of the structure and position of a target component provided in an embodiment of this application. The target cleaning distance between the drone 11 and the target component refers to the vertical distance (or height) between the drone 11 and the target component, for example... Figure 3 The target cleaning distance Hp is shown.

[0034] Step S300: Adjust the drone's flight altitude according to the target cleaning distance.

[0035] The target cleaning distance is greater than a preset safety threshold but less than a preset cleaning height. The preset cleaning height is the maximum cleaning height that the drone 11 can reach during cleaning operations above the target component; its specific value can be set according to actual conditions and is not specifically limited here. The preset safety threshold is a safety limit value that prevents the drone 11 from colliding with the target component during cleaning operations above it; its specific value can be set according to actual conditions and is not specifically limited here.

[0036] In the technical solution of this application embodiment, the specific implementation process of the flight altitude control method is as follows: First, image data and point cloud data collected in real time during the flight of the UAV 11 are acquired. Then, the target cleaning distance between the UAV 11 and the target component is determined in real time based on the image data and point cloud data. Since image data and point cloud data can complementaryly represent the positional information of the target component from both visual semantic and spatial geometric dimensions, the target cleaning distance determined based on the image data and point cloud data can accurately reflect the spatial positional relationship between the UAV 11 and the target component. Therefore, dynamically adjusting the flight altitude of the UAV 11 based on the target cleaning distance enables the UAV 11 to maintain a safe and effective working distance from the surface of the target component during the cleaning operation, thereby achieving precise adjustment of the flight altitude, which is beneficial to ensuring the cleaning effect and flight safety. Furthermore, this flight altitude control method does not need to consider the terrain scene where the target component is located, and has a wide range of applications.

[0037] In some embodiments, step S200 may include steps S210 to S230.

[0038] Step S210: Identify the target component based on the image data; The specific implementation method for identifying target components based on image data is as follows: If the image data contains part or all of the image of the target component, then it is determined that the target component has been identified at the current moment, indicating that the UAV 11 has flown above or near the target component. Otherwise, it is determined that the target component has not been identified, indicating that the UAV 11 has not yet flown above or near the target component.

[0039] Step S220: In response to the identification of the target component, determine the vertical distance between the UAV and the ground, and the component height of the target component based on the image data and point cloud data; The vertical distance between the drone 11 and the ground refers to the actual distance (or height) of the drone 11 above the ground in the vertical direction, for example... Figure 3 The height of the target component, as shown in the figure (Hg), is the actual distance (or height) of the center position of the target component's panel from the ground in the vertical direction. For example... Figure 3 The ΔH shown is given.

[0040] Step S230: Determine the target cleaning distance based on the vertical distance and component height.

[0041] The target cleaning distance is the difference between the vertical distance and the component height.

[0042] In the technical solution of this application embodiment, the specific implementation process of the flight altitude control method is as follows: First, image data and point cloud data collected in real time during the flight of the UAV 11 are acquired. Then, the target component is identified based on the image data to determine whether the target component has been identified. If part or all of the image of the target component is present in the image data, it is determined that the target component has been identified at the current moment, indicating that the UAV 11 has flown to the area above or near the target component. Further, the vertical distance between the UAV 11 and the ground, as well as the component height of the target component, are determined based on the image data and point cloud data. The target cleaning distance between the UAV 11 and the target component is accurately determined based on the vertical distance and component height, and then the flight altitude of the UAV 11 is adjusted based on the target cleaning distance. Thus, when the target component is identified, by linking the ground distance and the component height to calculate the target cleaning distance, the UAV 11 can directly use the surface of the component as the control target when it identifies the component, rather than using the ground or a preset flight path point as the control target. This reduces the problem of insufficient cleaning distance caused by the ground altitude fixation scheme not considering the height of the component itself in related technologies. Meanwhile, compared to the scheme of directly measuring the distance between components, the embodiment of this application introduces the vertical distance on the ground as an intermediate reference quantity, which provides a data basis for the flight control of the component-free area. This makes the entire cleaning flight control strategy logically consistent under different flight scenarios, which is conducive to improving the accuracy and effectiveness of the flight altitude control of UAV 11 and improving the reliability of UAV 11 cleaning operation.

[0043] In some embodiments, step S220 may include the following steps S221 to S223: Step S221: Determine the target component point cloud and ground point cloud based on the image data and point cloud data; The target component point cloud refers to the portion of the three-dimensional point cloud data acquired by the point cloud data sensor 122 during the flight of the UAV 11 that corresponds to the spatial location of the target component identified by the image data. The ground point cloud refers to the portion of the three-dimensional point cloud data acquired by the UAV 11 during the flight of the UAV 11 that is located below and around the target component and is used to characterize the ground elevation at the location of the target component.

[0044] Step S222: Determine the vertical distance based on the ground point cloud; Step S223: Determine the component height based on the target component point cloud and vertical distance.

[0045] Specifically, the target component point cloud and ground point cloud are determined from image data and point cloud data respectively. The vertical distance is determined based on the ground point cloud, and the component height is determined based on the target component point cloud and the vertical distance, thus achieving accurate calculation of the target cleaning distance. Since the component height is calculated by associating the component's spatial position reflected in the target component point cloud with the obtained ground vertical distance, and the vertical distance reflects the absolute height of the UAV 11 relative to the ground, while the component height reflects the relative height of the component, the difference between the two can accurately determine the actual working distance (i.e., the target cleaning distance) between the UAV 11 and the component. This improves the accuracy and effectiveness of the UAV 11's flight altitude control and enhances the reliability of the UAV 11's cleaning operation. Furthermore, in subsequent steps, outliers in any intermediate parameter (i.e., component height and vertical distance) can be detected and corrected, improving the robustness of the entire control system.

[0046] In some embodiments, step S221 may further include steps S2211 and S2212.

[0047] Step S2211: Fuse the image data and point cloud data, and find the point cloud in the point cloud data that corresponds to the target component in the image data as the target component point cloud; Specifically, the method for determining the target component point cloud can be as follows: First, the pixel region of the target component is detected in the image data using an image recognition algorithm. Then, using the spatial mapping relationship between the image and the point cloud (e.g., sensor calibration parameters), the three-dimensional spatial points corresponding to the target component region in the image are extracted from the original point cloud set to form the target component point cloud. This target component point cloud contains the three-dimensional geometric information of the target component surface, such as the spatial coordinates of the target component, the distribution of normal vectors, and the spatial range of the point cloud. It serves as the basic data source for subsequent operations such as plane fitting, tilt calculation, and edge height extraction, which is beneficial for further accurate determination of vertical distance and component height, thereby improving the accuracy of target cleaning height determination and the reliability of UAV cleaning operations.

[0048] Step S2212: Take the point cloud surrounding and below the target component in the point cloud data as the ground point cloud.

[0049] The point cloud surrounding and below the target component refers to the set of point cloud data located in the vicinity of the target component and below the target component in the three-dimensional point cloud data acquired during the flight of UAV 11.

[0050] Specifically, by fusing image data and point cloud data, the point cloud corresponding to the target component in the image is identified as the target component point cloud, and the point cloud surrounding and below the target component is used as the ground point cloud. This enables accurate alignment and association of the two heterogeneous data sets. Image data provides the semantic recognition result of the target component (i.e., its position in the image), while point cloud data provides the corresponding three-dimensional spatial coordinates. The resulting component point cloud, obtained through fusion, possesses both semantic accuracy and spatial precision. Furthermore, by using the point cloud surrounding and below the component as the ground point cloud, instead of globally searching the ground, the spatial range of ground detection is significantly reduced. This decreases the likelihood of misidentifying distant terrain undulations, shadow areas, or other irrelevant ground points as reference ground for the component's location, thereby improving the accuracy of vertical distance calculation and reducing the algorithm's computational complexity.

[0051] In some embodiments, step S222 may include steps S2221 to S2227.

[0052] Step S2221: Determine the component point cloud of the target component based on the point cloud data; Step S2222: Perform attitude calculation on the inertial measurement data during the flight of the UAV to obtain quaternions; wherein, the inertial measurement data is obtained by the IMU sensor 123 in real time.

[0053] Step S2223: Perform coordinate transformation on the quaternion to obtain the rotation matrix of the inertial measurement unit coordinate system relative to the world coordinate system; Step S2224: Perform an inverse transformation on the rotation matrix to obtain the inverse matrix; Step S2225: Perform coordinate transformation on the point cloud data using an inverse matrix, and determine the lower part of the component point cloud based on all the point cloud coordinate values ​​after coordinate transformation; Step S2226: Use the point cloud below as the ground point cloud; Step S2227: Average the ground point cloud to obtain the vertical distance.

[0054] For example, the vertical distance between the UAV 11 and the ground can be obtained by processing the ground point cloud (e.g., taking an average value), such as... Figure 3 The vertical distance Hg is shown.

[0055] Figure 4 This is a top view of the target component provided in an embodiment of this application. For example, see [link to example]. Figure 4 Q1 is the point cloud region of the target component, and Q2 is the point cloud region below and around the target component (i.e., the ground point cloud).

[0056] For example, the process of determining the ground point cloud below and around a target component is as follows: First, attitude calculation is performed on the IMU sensor data (accelerometer and angular velocity data) to obtain quaternions, for example, using at least one of the Madgwick algorithm, Mahony algorithm, and Kalman filtering. Then, the rotation matrix of the IMU coordinate system relative to the world coordinate system is obtained from the quaternions. Next, the inverse matrix of the rotation matrix is ​​obtained to obtain the inverse transformation from the world coordinate system to the IMU coordinate system. Finally, the point cloud data is transformed using this inverse transformation matrix. After the transformation, the Z-direction of all point cloud coordinate values ​​is downward, and the point cloud below is the ground point cloud.

[0057] Therefore, by obtaining quaternions through IMU attitude calculation, a rotation matrix and its inverse matrix are obtained. The inverse matrix is ​​then used to perform coordinate transformation on the point cloud data, thus identifying the point cloud located "below" as the ground point cloud. Finally, the vertical distance is obtained by averaging this ground point cloud. Because the gravity direction provided by the IMU ensures the stability of the "below" determination under an absolute reference frame, the Z-axis of the point cloud after coordinate transformation always points towards the Earth's center, regardless of whether the UAV 11 experiences roll, pitch, or other attitude changes during flight. This ensures that the selected ground point cloud always represents the ground directly below the UAV in the physical world, reducing ranging errors caused by the UAV 11's own attitude tilt and achieving consistency in vertical distance across different flight attitudes. This is particularly beneficial for improving reliability in dynamic environments such as wind disturbances. Furthermore, by using the point cloud surrounding the components as the ground, the complexity and computational load of ground detection are reduced, thereby improving the reliability of calculating the component's ground altitude.

[0058] In some embodiments, step S223 may further include steps S2231 to S2233.

[0059] Step S2231: Determine the target parameters of the target component based on the point cloud of the target component; The target parameters include the tilt angle of the target component and the distance between the edge of the target component and the UAV 11 (referred to as edge distance).

[0060] Figure 5 This is a schematic diagram illustrating the target parameters of a target component provided in an embodiment of this application. The tilt angle of the target component refers to the spatial tilt angle of the photovoltaic module's panel plane relative to the horizontal plane (or the direction of gravity), for example... Figure 5 The angle 'a' or 'b' shown represents the orientation of the photovoltaic module in three-dimensional space. Edge distance refers to the height (or distance) of the four boundaries of the photovoltaic module panel from the drone in the direction of gravity (vertical direction), for example... Figure 5 The distance from the first edge of one boundary of the target component is H1, and the distance from the second edge of the other boundary of the target component is H2.

[0061] Step S2232: Determine the current actual distance between UAV 11 and the target component based on the target parameters; Step S2233: Determine the component height based on the difference between the vertical distance and the current actual distance.

[0062] Specifically, the target parameters of the target component (including tilt and the distance between the edge and the UAV) are first determined based on the target component point cloud. Then, the current actual distance between the UAV 11 and the target component is determined based on these target parameters. Finally, the component height is determined by the difference between the vertical distance and the current actual distance. Since the tilt in the target parameters reflects the spatial orientation of the target component, and the edge distance reflects the specific position of each boundary of the target component in space, the current actual distance determined based on these two factors can represent the overall spatial position of the target component rather than the position of a single point on the component. Therefore, the component height can be accurately obtained by the difference between the current actual distance and the vertical distance, which can reflect the true ground clearance of the target component. Furthermore, using this component height as the data source for subsequent virtual height estimation has higher reliability and representativeness, providing an accurate data foundation for flight control in component-free areas. This is beneficial for improving the accuracy and effectiveness of UAV 11's flight altitude control and the reliability of UAV 11's cleaning operations.

[0063] In some embodiments, step S2231 may include steps S22311 to S22316.

[0064] Step S22311: Perform planar fitting on the point cloud of the target component and extract the target point cloud that is on the same fitting reference plane; Among them, the fitting reference surface is used to characterize the spatial orientation of the target component; Figure 6 This is a schematic diagram of the fitting reference plane provided in an embodiment of this application. Since the target component point cloud is randomly distributed, by performing planar fitting on the target component point cloud, target point clouds on the same plane can be fitted to the same fitting reference plane, and non-planar point clouds are filtered out, resulting in... Figure 6 The fitted reference plane P0 is shown, and the normal vector F0 or equation coefficients of the fitted reference plane P0 can be obtained simultaneously. Among them, the equation coefficients refer to the x, y, and z coefficients of the general equation of the three-dimensional plane (Ax+By+Cz+D=0). The normal vector can be directly constructed through the x, y, and z coefficients. For example, the vector n=(A,B,C) is a normal vector of the plane.

[0065] Step S22312: Determine the normal vector of the fitting reference surface; Step S22313: Perform spatial bounding box processing on the fitted reference plane to generate the spatial envelope of the target point cloud; The spatial bounding box processing can be performed using an Oriented Bounding Box (OBB). The spatial envelope can be a three-dimensional rectangle.

[0066] Figure 7 This is a schematic diagram of a spatial envelope provided in an embodiment of this application. For example, by performing spatial bounding box processing on the target point cloud within the fitting reference plane, a spatial envelope capable of enclosing the target point cloud is generated, such as... Figure 7 The figure shows the minimum bounding 3D rectangle P1 that can contain the target point cloud.

[0067] Step S22314: Based on the edge orientation information of the spatial envelope, extract a two-dimensional rectangular frame that is coplanar with the fitting reference plane inside the envelope; The edge orientation information of the spatial envelope includes the shortest edge of the spatial envelope, or the four edges most parallel to the direction of the normal vector F0, for example... Figure 7 The minimum enclosing three-dimensional rectangle P1 is shown with its first vertical edge L1, second vertical edge L2, third vertical edge L3, and fourth vertical edge L4.

[0068] Since the 3D rectangle P1 is the smallest envelope that can enclose the target point cloud, and its volume and shape are basically the same as the fitting reference plane, a 2D rectangle coplanar with the fitting reference plane is extracted inside the envelope. For example, the 2D rectangle obtained by intersecting the 3D rectangle P1 with the plane formed by the midpoints of the four shortest sides of the 3D rectangle P1 (e.g., the first vertical edge L1, the second vertical edge L2, the third vertical edge L3, and the fourth vertical edge L4 of the 3D rectangle P1) or the four sides most parallel to the direction of the normal vector F0 is used as the ideal planar rectangle of the planar point cloud (i.e., the target point cloud).

[0069] Wherein, the two-dimensional rectangular frame is an ideal planar rectangle, for example Figure 7 The given diagram shows a two-dimensional rectangular frame composed of points A, B, C, and D. The midpoint of the two-dimensional rectangular frame can be obtained through the four vertices of an ideal planar rectangle (e.g., points A, B, C, and D), and the equation parameters of the four edges can also be obtained.

[0070] Step S22315: Determine the tilt angle based on the normal vector and the inertial measurement data during the flight of UAV 11; The direction of gravity can be calculated using inertial measurement data, since the fitting reference plane is the same as the normal vector F0 of the two-dimensional rectangular frame. The tilt of the target component can be obtained by calculating the angle between the normal vector F0 and the direction of gravity.

[0071] Step S22316: Determine the distance between the edge of the target component and the drone 11 based on the average distance from all points on each side of the two-dimensional rectangle to the drone 11.

[0072] Among them, the average value of each side has an average maximum value and an average minimum value. The average maximum value is used as the height of the two-dimensional rectangle (i.e., the first edge distance), and the average minimum value is used as the low side of the two-dimensional rectangle (i.e., the second edge distance).

[0073] Specifically, select some points from the four sides of the two-dimensional rectangle, for example, in Figure 7 We take some points on edge AB and calculate an average value, take some points on edge BC and calculate an average value, take some points on edge CD and calculate an average value, and take some points on edge AD and calculate an average value. The highest average value among these four values ​​is taken as the high edge, and the lowest average value among these four values ​​is taken as the low edge. In this way, we can accurately calculate the height of the two edges of the target component (i.e., the distance between the first edge and the distance between the second edge), which is beneficial for accurately calculating the target cleaning distance and improving the accuracy of flight altitude control.

[0074] In some embodiments, the distance between the edge of the target component and the drone includes a first edge distance and a second edge distance; the first edge distance is greater than the second edge distance. The above step S2232 may include the following steps S22321 and S22322.

[0075] Step S22321: Determine the height center position parameter of the target component based on the geometric relationship between the first edge distance, the second edge distance, and the tilt. The height center position parameter includes the midpoint between the first edge distance and the second edge distance, or the average of the first edge distance and the second edge distance. The first edge distance refers to the vertical distance from the UAV 11 to the edge with the largest average distance value among the four boundary edges of the ideal planar rectangle of the target component. The second edge distance refers to the vertical distance from the UAV 11 to the edge with the smallest average distance value among the four boundary edges of the ideal planar rectangle of the target component.

[0076] Furthermore, the height center position parameter can also be the average of multiple edge distance values ​​between the first edge distance and the second edge distance. It can be the median of multiple edge distances between the first edge distance and the second edge distance, or it can be the first edge distance, the second edge distance, and the average of multiple edge distances between the first edge distance and the second edge distance. For example, based on the trapezoidal geometry formed by the first edge distance H1, the inclination (e.g., angle b), the second edge distance H2, and the horizontal line x, multiple edge distance values ​​between the first edge distance H1 and the second edge distance H2 can be obtained using mathematical calculation algorithms or formulas. Figure 5 The distances to the third edge (H3), fourth edge (H4), and fifth edge (H5) are calculated. Then, the average of these edge distances is used as the height center position parameter. Alternatively, the median of the edge distances can be used; the specific setting depends on the actual situation and is not limited here.

[0077] Since photovoltaic modules are typically tilted, the distances between the high and low sides and the drone 11 (or sensor 12) are different. For ease of calculation, all subsequent heights or distances related to the module will be represented by intermediate values. Knowing the tilt angle, the heights on both sides can be calculated. For example, this embodiment uses the intermediate value as the height center position parameter for illustration.

[0078] Step S22322: Determine the current actual distance based on the height center position parameters.

[0079] Specifically, after determining the tilt angle and edge distance of the target component, the height center position parameter of the target component (such as the aforementioned intermediate or average value) can be obtained based on the geometric relationship between the first edge distance, the second edge distance, and the tilt angle. This height center position parameter is then used as the current actual distance between the UAV 11 and the target component. Since the tilt angle of the target component and the edge distance between the target component and the UAV 11 (i.e., the height center position parameter) accurately reflect the pose of the target component in space and its spatial positional relationship with the UAV, determining the current actual distance based on the height center position parameter and further calculating the component height precisely based on the difference between the vertical distance and the current actual distance is beneficial for accurately adjusting the flight altitude of the UAV 11. This allows the UAV 11 to maintain a safe and effective working distance from the target component surface during cleaning operations, thereby achieving precise adjustment of the flight altitude.

[0080] In some embodiments, step S200 may further include steps S201 to S204.

[0081] Step S201: In response to the failure to identify the target component based on the image data, determine the neighboring components within the preset spatial neighborhood of the location of the UAV 11, and use the neighboring components as the target components; Step S202: Obtain historical component height records of neighboring components; Step S203: Determine the vertical distance between the UAV and the ground based on the image data and point cloud data; Step S204: Determine the target cleaning distance based on the vertical distance and historical component height records.

[0082] The preset spatial neighborhood refers to the three-dimensional spatial range defined by extending outwards from the spatial location of the target component as the center or reference along the horizontal and / or vertical directions by a preset distance threshold. The preset distance threshold can be pre-set based on the arrangement spacing of photovoltaic modules in the photovoltaic power station, the positioning accuracy of the UAV 11, and the flight speed of the cleaning operation. For example, the preset distance threshold can be set to 3 meters, 5 meters, 10 meters, 20 meters, etc.

[0083] Historical component altitude record refers to the data record formed by the UAV 11 as it associates and stores the spatial location information of the photovoltaic component with its altitude after it identifies the photovoltaic component and calculates its altitude above the ground during flight.

[0084] The specific implementation process of determining the vertical distance between the UAV and the ground based on image data and point cloud data in step S203 is the same as that in steps S2221 to S2227 above, and will not be repeated here.

[0085] Specifically, if the target component is not identified, the ground is identified, and the vertical distance between the UAV 11 (or sensor 12) and the ground is calculated based on point cloud data. Since the flight altitude of the UAV 11 is adjusted relative to the component, and no component is identified at this time, only the vertical distance is available without the component height ΔH of the target component, so the flight altitude cannot be adjusted. Therefore, by acquiring historical component height records within a preset spatial neighborhood of the target component's location, and determining the target cleaning distance based on whether such records exist, a strategy for determining the target cleaning distance is established. If historical component height records exist, the target cleaning distance of the target component is determined based on the historical component height records and the vertical distance. This allows the UAV 11, when flying into areas with component gaps or where image recognition fails, to no longer attempt to track the ground and dive downwards, thus avoiding the risk of collision, as is done with related technologies that use panel-based altitude hold schemes. Instead, it uses historical height data of neighboring components within the spatial neighborhood to infer the appropriate component height at the current location, thereby maintaining a safe and reasonable flight altitude. At the same time, the spatial constraints of the preset spatial neighborhood ensure that the referenced historical data has spatial correlation with the current location, improving the effectiveness of virtual altitude estimation.

[0086] In some embodiments, step S204 may further include steps S2041 and S2042.

[0087] Step S2041: In response to the existence of historical component height records, determine the virtual height of the target component based on the historical component height records; If a historical component height record exists, the virtual height of the target component is determined based on that record. For example, based on the height values ​​of several photovoltaic modules surrounding the target module closest to the current location (i.e., the location of the target module), H) Calculate the virtual height of the target component. Specifically, the height values ​​of each component can be weighted according to the distance of surrounding components from the target component (e.g., the closer the surrounding components are to the target component, the higher their weight), and the average value can be calculated and used as the virtual height. .

[0088] Step S2042: Determine the target cleaning distance based on the difference between the vertical distance and the virtual height.

[0089] Specifically, when historical component height records exist, the virtual height of the target component is determined based on these records, and the target cleaning distance is determined by the difference between the vertical distance and the virtual height. Since the ground clearance of photovoltaic components is spatially continuous (the installation height of adjacent components typically changes gradually), using the actual height records of neighboring components to estimate the virtual component height at the current location is physically reasonable. Therefore, in areas without components, the virtual component height (i.e., the virtual height) is determined based on the ground clearance of multiple nearby components. The system calculates the target cleaning distance based on the virtual component height and vertical distance to adjust the flight altitude, thereby improving flight smoothness, reducing the risk of drone collisions, improving the accuracy and effectiveness of drone 11 flight altitude control, and improving the reliability of drone 11 cleaning operations.

[0090] In some embodiments, the flight altitude control method further includes: in response to the absence of historical component altitude records, determining the target cleaning distance based on a preset safe flight altitude.

[0091] The preset safe flight altitude refers to the safe cleaning flight altitude at which the drone 11 will not collide with the photovoltaic modules. Its specific value can be set according to the actual situation and is not specifically limited here.

[0092] Specifically, when the ground is detected and no historical component height records exist, a preset safe flight altitude is used as the target cleaning distance to control the drone 11 to fly at the preset safe flight altitude. For example, when the drone 11 first takes off, it flies along a pre-planned path, which is usually a latitude and longitude coordinate system and does not include ground altitude information. The takeoff point typically does not have photovoltaic components, so the components cannot be detected at takeoff; only the ground is detected. Since there are no component height records at takeoff, no historical component height records exist, and altitude adjustments cannot be made. Therefore, the drone 11 is initially controlled to fly at the preset safe flight altitude. As the number of operations increases, the historical component height records become richer, and the accuracy of virtual altitude estimation continuously improves, making the drone 11's flight trajectory smoother and improving the reliability of flight control and cleaning operations.

[0093] Figure 8 This is a schematic diagram of the overall process of a flight altitude control method provided in an embodiment of this application. For example, taking the target cleaning distance as the vertical distance between the sensor and the component, and the distance between the sensor and the ground as an example, see [reference needed]. Figure 8 The overall process of this flight altitude control method includes the following steps: Step S600: After the UAV 11 takes off, it flies along the planned path.

[0094] Step S601: Simultaneously collect sensor data and provide it to the computing power box 13, which then identifies the components.

[0095] Step S602: Determine whether the component is identified. If yes, proceed to step S610; otherwise, proceed to step S620.

[0096] Step S610: Calculate the vertical distance Hp between the component panel and the sensor.

[0097] Step S611: Using the point cloud below the component as the ground, calculate the vertical distance Hg between the sensor and the ground.

[0098] Step S612: Use the difference ΔH between Hp and Hg as the component height, and record the height and position of each component.

[0099] Step S613: Adjust the flight altitude according to the component height ΔH and the current ground distance (i.e., the vertical distance between the sensor and the ground) Hg.

[0100] Step S620: Identify the ground and calculate the vertical distance Hg between the sensor and the ground.

[0101] Step S621: Determine if there is a historical component height ΔH record. If there is, proceed to step S622; otherwise, proceed to step S623.

[0102] Step S622: Calculate the height of the virtual component at the current position based on the nearest set of ΔH values. (i.e., virtual height).

[0103] Step S23: Adjust the flight altitude according to the preset ground clearance (i.e., the preset safe flight altitude).

[0104] Step S624: According to the height of the virtual component Adjust the flight altitude based on the current ground distance Hg.

[0105] Figure 9 This is a schematic block diagram of a flight altitude control device provided in an embodiment of this application. This application also provides a flight altitude control device, see below. Figure 9 The flight altitude control device 20 includes: an acquisition module 21 for acquiring image data and point cloud data collected during the flight of the UAV 11; a determination module 22 for determining the target cleaning distance between the UAV 11 and the target component based on the image data and point cloud data; and an adjustment module 23 for adjusting the flight altitude of the UAV 11 based on the target cleaning distance.

[0106] The technical solution of this application provides a flight altitude control device that acquires real-time image data and point cloud data collected during the flight of a drone, and determines the target cleaning distance between the drone and the target component in real time based on the image data and point cloud data. Since image data and point cloud data can complementaryly represent the positional information of the target component from both visual semantic and spatial geometric dimensions, the target cleaning distance determined based on the image data and point cloud data can accurately reflect the spatial positional relationship between the drone and the target component. Therefore, dynamically adjusting the drone's flight altitude based on the target cleaning distance enables the drone to maintain a safe and effective working distance from the target component surface during cleaning operations, thereby achieving precise adjustment of the flight altitude.

[0107] In some embodiments, the determining module 22 is further configured to: Identify target components based on image data; In response to the identification of the target component, the vertical distance between the UAV and the ground, as well as the component height of the target component, are determined based on image data and point cloud data; The target cleaning distance is determined based on the vertical distance and component height.

[0108] In some embodiments, the determining module 22 is further configured to: The target component point cloud and ground point cloud are determined based on image data and point cloud data; Determine the vertical distance based on the ground point cloud; The component height is determined based on the target component point cloud and vertical distance.

[0109] In some embodiments, the determining module 22 is further configured to: Image data and point cloud data are fused together, and the point cloud corresponding to the target component in the image data is found in the point cloud data as the target component point cloud; The point cloud surrounding and below the target component in the point cloud data is used as the ground point cloud.

[0110] In some embodiments, the determining module 22 is further configured to: Determine the component point cloud of the target component based on the point cloud data; Quaternions are obtained by performing attitude calculations on inertial measurement data during the flight of the UAV. By performing coordinate transformation on the quaternion, the rotation matrix of the inertial measurement unit coordinate system relative to the world coordinate system is obtained; Perform an inverse transformation on the rotation matrix to obtain the inverse matrix; The coordinates of the point cloud data are transformed using an inverse matrix, and the bottom of the component point cloud is determined based on all the point cloud coordinate values ​​after the coordinate transformation. Use the point cloud below as the ground point cloud; The vertical distance is obtained by averaging the ground point cloud.

[0111] In some embodiments, the determining module 22 is further configured to: The target parameters of the target component are determined based on the point cloud of the target component; the target parameters include the tilt angle of the target component and the distance between the edge of the target component and the UAV. Determine the current actual distance between the UAV and the target components based on the target parameters; The component height is determined based on the difference between the vertical distance and the current actual distance.

[0112] In some embodiments, the determining module 22 is further configured to: Planar fitting is performed on the point cloud of the target component to extract the target point cloud that is located on the same fitting reference plane; wherein, the fitting reference plane is used to characterize the spatial orientation of the target component; Determine the normal vector of the fitting reference surface; The fitting reference surface is processed by a spatial bounding box to generate a spatial envelope of the target point cloud. Based on the edge orientation information of the spatial envelope, a two-dimensional rectangular frame coplanar with the fitting reference plane is extracted inside the envelope; The tilt angle is determined based on the normal vector and inertial measurement data during the UAV's flight. The distance between the edge of the target component and the drone is determined by the average distance from all points on each side of the two-dimensional rectangle to the drone.

[0113] In some embodiments, the distance between the edge of the target component and the drone includes a first edge distance and a second edge distance; the first edge distance is greater than the second edge distance. Module 22 is also used for: Based on the geometric relationship between the first edge distance, the second edge distance, and the tilt, the height center position parameter of the target component is determined; the height center position parameter includes the median value of the first edge distance and the second edge distance, or the average value of the first edge distance and the second edge distance; The current actual distance is determined based on the height center position parameter.

[0114] In some embodiments, the determining module 22 is further configured to: In response to the failure to identify the target component based on the image data, the neighboring components within a preset spatial neighborhood of the drone's location are determined and the neighboring components are used as the target components. Retrieve historical component height records from neighboring components; Based on image data and point cloud data, determine the vertical distance between the drone and the ground; The target cleaning distance is determined based on the vertical distance and historical component height records.

[0115] In some embodiments, the determining module 22 is further configured to: In response to the existence of historical component height records, the virtual height of the target component is determined based on the historical component height records; The target cleaning distance is determined based on the difference between the vertical distance and the virtual height.

[0116] In some embodiments, the flight altitude control device 20 further includes a determining unit for: In response to the absence of historical component altitude records, the target cleaning distance is determined based on the preset safe flight altitude.

[0117] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0118] The foregoing has provided a detailed description of a flight altitude control method, apparatus, and cleaning system provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for controlling flight altitude, characterized in that, A drone used for cleaning photovoltaic modules; the method includes: Acquire image data and point cloud data collected during the flight of the UAV; Based on the image data and the point cloud data, the target cleaning distance between the UAV and the target component is determined; The drone's flight altitude is adjusted according to the target cleaning distance.

2. The method according to claim 1, characterized in that, Determining the target cleaning distance between the UAV and the target component based on the image data and the point cloud data includes: Identify the target component based on the image data; In response to the identification of the target component, the vertical distance between the UAV and the ground, and the component height of the target component are determined based on the image data and the point cloud data. The target cleaning distance is determined based on the vertical distance and the component height.

3. The method according to claim 2, characterized in that, Determining the vertical distance between the UAV and the ground, and the component height of the target component, based on the image data and the point cloud data, includes: The target component point cloud and the ground point cloud are determined based on the image data and the point cloud data; The vertical distance is determined based on the ground point cloud; The height of the component is determined based on the point cloud of the target component and the vertical distance.

4. The method according to claim 3, characterized in that, The step of determining the target component point cloud and ground point cloud based on the image data and the point cloud data includes: The image data and the point cloud data are fused together, and the point cloud data that corresponds to the target component in the image data is found in the point cloud data as the target component point cloud; The point cloud surrounding and below the target component in the point cloud data is taken as the ground point cloud.

5. The method according to claim 3, characterized in that, The method for determining the vertical distance includes: The component point cloud of the target component is determined based on the point cloud data; The attitude calculation of the inertial measurement data during the flight of the UAV yields quaternions; The quaternion is transformed to obtain the rotation matrix of the inertial measurement unit coordinate system relative to the world coordinate system. Perform an inverse transformation on the rotation matrix to obtain the inverse matrix; The inverse matrix is ​​used to perform coordinate transformation on the point cloud data, and the lower part of the component point cloud is determined based on all the point cloud coordinate values ​​after the coordinate transformation. The point cloud below is taken as the ground point cloud; The vertical distance is obtained by averaging the ground point cloud.

6. The method according to claim 3, characterized in that, Determining the component height based on the target component point cloud and the vertical distance includes: The target parameters of the target component are determined based on the point cloud of the target component; the target parameters include the tilt angle of the target component and the distance between the edge of the target component and the UAV; The current actual distance between the UAV and the target component is determined based on the target parameters; The component height is determined based on the difference between the vertical distance and the current actual distance.

7. The method according to claim 6, characterized in that, Determining the target parameters of the target component based on the point cloud of the target component includes: Planar fitting is performed on the point cloud of the target component to extract the target point cloud located on the same fitting reference plane; wherein, the fitting reference plane is used to characterize the spatial orientation of the target component; Determine the normal vector of the fitted reference surface; The fitting reference surface is subjected to spatial bounding box processing to generate a spatial envelope that encloses the target point cloud; Based on the edge orientation information of the spatial envelope, a two-dimensional rectangular frame coplanar with the fitting reference plane is extracted inside the envelope; The tilt angle is determined based on the normal vector and the inertial measurement data during the flight of the UAV; The distance between the edge of the target component and the drone is determined based on the average distance from all points on each side of the two-dimensional rectangle to the drone.

8. The method according to claim 7, characterized in that, The distance between the edge of the target component and the drone includes a first edge distance and a second edge distance; the first edge distance is greater than the second edge distance. Determining the current actual distance between the UAV and the target component based on the target parameters includes: The height center position parameter of the target component is determined based on the geometric relationship between the first edge distance, the second edge distance, and the tilt angle. The height center position parameter includes the median value of the first edge distance and the second edge distance, or the average value of the first edge distance and the second edge distance; The current actual distance is determined based on the height center position parameter.

9. The method according to claim 1, characterized in that, Determining the target cleaning distance between the UAV and the target component based on the image data and the point cloud data includes: In response to the failure to identify the target component based on the image data, a neighboring component within a preset spatial neighborhood of the location of the drone is determined, and the neighboring component is used as the target component; Obtain the historical component height records of the neighboring components; Based on the image data and the point cloud data, the vertical distance between the UAV and the ground is determined; The target cleaning distance is determined based on the vertical distance and the historical component height records.

10. The method according to claim 9, characterized in that, Determining the target cleaning distance based on the vertical distance and the historical component height records includes: In response to the existence of the historical component height record, the virtual height of the target component is determined based on the historical component height record; The target cleaning distance is determined based on the difference between the vertical distance and the virtual height.

11. The method according to claim 9, characterized in that, The method further includes: in response to the absence of the historical component altitude record, determining the target cleaning distance based on a preset safe flight altitude.

12. A flight altitude control device, characterized in that, A drone used for cleaning photovoltaic modules; the device includes: The acquisition module is used to acquire image data and point cloud data collected during the flight of the UAV; The determination module is used to determine the target cleaning distance between the UAV and the target component based on the image data and the point cloud data; An adjustment module is used to adjust the flight altitude of the UAV according to the target cleaning distance.

13. A cleaning system for photovoltaic modules, characterized in that, include: Drones used for cleaning photovoltaic modules; Sensors mounted on the underside of the drone are used to collect image data and point cloud data during the drone's flight. The computing power box is used to determine the target cleaning distance between the drone and the target component based on the image data and the point cloud data; The control module is used to adjust the flight altitude of the UAV according to the target cleaning distance.

Images