A method and apparatus for controlling the operation of a photovoltaic cleaning machine.
By constructing a point cloud map of the photovoltaic sweeper using lidar and ranging sensors, the slope of the bridge can be identified and the speed adjusted, solving the problem of the photovoltaic sweeper crossing in complex environments and improving the stability and lifespan of the equipment.
Patent Information
- Application Number
- CN202211673870.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-26
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2042-12-26
AI Technical Summary
When faced with complex environments, such as steep and misaligned cable trays, photovoltaic sweepers have difficulty crossing smoothly and are prone to jamming, tipping over, or wear and tear. They are also prone to violent collisions at docking or return stations.
A point cloud map of the cleaning environment is constructed by a lidar installed on the photovoltaic sweeper to identify the cable tray and obtain slope information. Combined with distance measurement sensors, the speed of the photovoltaic sweeper is adjusted to avoid jamming and tipping over. Ultrasonic radar and motor encoders are used to control deceleration and reduce the risk of wear.
This enabled the photovoltaic sweeper to successfully navigate complex environments, avoiding jamming, tipping over, and wear, thus improving the equipment's operational stability and lifespan.
Smart Images

Figure CN115940785B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of photovoltaic cleaning, and in particular to a method and apparatus for controlling the operation of a photovoltaic cleaning machine. Background Technology
[0002] The photovoltaic cleaning robot utilizes the rapid rotation of its brushes to clean along the photovoltaic modules. Besides ensuring smooth movement along the modules, it must also be able to traverse different support structures, moving from one photovoltaic support to another. Normally, bridges need to be erected between supports with large angles and gaps to allow the robot to cross smoothly. However, due to the complex environment, sometimes the slope of these bridges is too steep, preventing the robot from climbing the modules at normal operating speed. Alternatively, excessive angles can cause the robot to get stuck or tip over at normal speeds. When approaching or returning to a stopping point, if the robot doesn't slow down, it is prone to frequent and violent collisions, leading to severe wear on its wheels. Summary of the Invention
[0003] To address the above problems, the purpose of this invention is to provide a method and device for controlling the operation of a photovoltaic cleaning machine, enabling the photovoltaic cleaning machine to successfully cross complex environments (such as steep and misaligned cable trays), avoiding problems such as being unable to climb photovoltaic modules due to speed issues, or getting stuck or tipping over due to excessive slope angles.
[0004] The technical solution provided by this invention is a method for controlling the operation of a photovoltaic cleaning machine, as follows:
[0005] In some implementations, the steps include:
[0006] A point cloud map of the cleaning environment is constructed using a lidar installed on a photovoltaic cleaning machine;
[0007] Based on the point cloud map of the cleaning environment, the cable tray in front is identified, and the slope information of the cable tray is obtained;
[0008] When the cable tray is detected, the current distance to the cable tray is measured by a distance sensor installed on the photovoltaic cleaning machine;
[0009] The speed of the photovoltaic sweeper is adjusted based on the current distance to the cable tray and the slope information of the cable tray.
[0010] In some embodiments, the construction of a point cloud map of the cleaning environment using a lidar installed on a photovoltaic cleaning machine specifically includes:
[0011] At least two environmental point clouds are acquired by scanning with at least two lidar sensors installed on the photovoltaic cleaning machine; the two lidar sensors are on the same horizontal or vertical line.
[0012] Extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud;
[0013] Voxel filtering is used to downsample the truncated effective point cloud;
[0014] Set up NDT (Normal Distributions Transform), calculate the transformation matrix, and perform NDT registration on the downsampled at least two environmental point clouds based on the transformation matrix;
[0015] The registered at least two environmental point clouds are stitched together to generate a 3D cleaning environment point cloud map.
[0016] In some implementations, the step of identifying the cable tray ahead based on the point cloud map of the cleaning environment and obtaining the slope information of the cable tray specifically includes:
[0017] Based on the cleaning environment point cloud map, photovoltaic modules and cable trays in the cleaning environment point cloud map are identified;
[0018] The point clouds of the photovoltaic modules and cable trays were segmented using a semantic segmentation algorithm;
[0019] Project the point clouds of the photovoltaic modules and cable trays to obtain a 2D target projection point cloud map;
[0020] Linearity fitting is performed on the point cloud in the 2D target projection point cloud map to obtain the linear expressions for the photovoltaic module and the bridge.
[0021] Based on the linear expressions of the photovoltaic module and the cable tray, the relative angle between the cable tray and the photovoltaic module is determined.
[0022] In some embodiments, obtaining the slope information of the cable tray specifically includes:
[0023] The first distance and the second distance to the front cable tray are measured by the first distance sensor and the second distance sensor installed on the same vertical line on the photovoltaic sweeper, respectively;
[0024] Obtain the difference between the first distance and the second distance;
[0025] Based on the difference, and combined with the installation distance between the first and second ranging sensors, the slope of the cable tray is obtained.
[0026] In some embodiments, adjusting the speed of the photovoltaic sweeper based on the current distance to the cable tray and the slope information of the cable tray specifically includes:
[0027] Based on the slope information of the cable tray, the target speed of the photovoltaic sweeper when it reaches the cable tray is found from the pre-stored slope-speed correspondence table;
[0028] Based on the current distance to the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper, the corresponding acceleration is obtained;
[0029] The operation of the photovoltaic sweeper is controlled according to the acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
[0030] In some embodiments, the operation control method for the photovoltaic sweeper further includes:
[0031] Based on the point cloud map of the cleaning environment, identify the stop or return station ahead;
[0032] When the stop or return station is detected, the distance between the photovoltaic sweeper and the stop or return station is measured by a distance sensor installed on the photovoltaic sweeper.
[0033] Based on the distance between the photovoltaic sweeper and the docking station or return station, and combined with the current speed of the photovoltaic sweeper, the speed of the photovoltaic sweeper is reduced so that the speed at which the photovoltaic sweeper arrives at the docking station or return station is less than a preset speed.
[0034] In some embodiments, based on the same technical concept, this solution also provides a photovoltaic cleaning and operation control device, including:
[0035] The point cloud map building module is used to build a point cloud map of the cleaning environment using a lidar installed on a photovoltaic sweeper.
[0036] The slope information acquisition module is used to identify the cable tray in front based on the point cloud map of the cleaning environment and to acquire the slope information of the cable tray;
[0037] A distance measurement module is used to measure the current distance to the cable tray when the cable tray is detected, using a distance sensor installed on the photovoltaic cleaning machine.
[0038] The speed adjustment module is used to adjust the speed of the photovoltaic sweeper based on the current distance to the cable tray and the slope information of the cable tray.
[0039] In some implementations, the point cloud map construction module specifically includes:
[0040] An environmental point cloud acquisition submodule is used to acquire at least two environmental point clouds by scanning with at least two lidars installed on the photovoltaic cleaning machine; the two lidars are on the same horizontal or vertical line.
[0041] The noise filtering submodule is used to extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud.
[0042] The downsampling submodule is used to downsample the truncated effective point cloud using voxel filtering.
[0043] The point cloud map construction submodule is used to set NDT, calculate the transformation matrix, and perform NDT registration on the downsampled at least two environmental point clouds based on the transformation matrix; the registered at least two environmental point clouds are then stitched together to generate a 3D cleaning environment point cloud map.
[0044] In some implementations, the slope information acquisition module specifically includes:
[0045] The identification submodule is used to identify photovoltaic modules and cable trays in the cleaning environment point cloud map based on the cleaning environment point cloud map;
[0046] The algorithm segmentation submodule is used to segment the point cloud of the photovoltaic module and the cable tray using a semantic segmentation algorithm;
[0047] The projection submodule is used to project the point cloud of the photovoltaic module and the cable tray to obtain a 2D target projection point cloud map;
[0048] The fitting submodule is used to linearize and fit the point cloud in the 2D target projection point cloud map to obtain the linear expression of the photovoltaic module and the bridge.
[0049] The slope acquisition submodule is used to determine the relative angle between the photovoltaic module and the photovoltaic module based on the linear expression of the photovoltaic module and the cable tray.
[0050] In some embodiments, the speed adjustment module specifically includes:
[0051] The speed acquisition submodule is used to look up the target speed of the photovoltaic sweeper when it reaches the bridge from a pre-stored slope-speed correspondence table based on the slope information of the bridge.
[0052] An acceleration acquisition submodule is used to obtain the corresponding acceleration based on the current distance to the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper;
[0053] The speed adjustment submodule is used to control the operation of the photovoltaic sweeper according to the acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
[0054] The operation control method and apparatus for a photovoltaic cleaning machine provided by this invention have at least the following features:
[0055] Beneficial effects:
[0056] 1. In this invention, for complex scenarios with large slopes such as cable trays with large slopes or large misalignments, the slope of the cable tray ahead can be identified in a timely manner and the slope of the cable tray ahead can be calculated. Then, the running speed of the photovoltaic cleaning machine can be controlled and adjusted so that it can cross the cable tray and reach another photovoltaic support to be cleaned smoothly. This avoids the phenomenon that the photovoltaic cleaning machine cannot climb the connecting cable tray at the current speed and is prone to getting stuck or tipping over.
[0057] 2. The operation control method and device for a photovoltaic sweeper provided by this invention employs a photovoltaic sweeper body equipped with a lidar + ultrasonic radar or infrared sensor. The lidar can pre-map the entire operating route, obtaining a point cloud map. This facilitates early identification of locations along the route that may cause operational malfunctions, allowing for proactive prediction.
[0058] 3. The operation control method and device of the photovoltaic sweeper provided by the present invention utilizes the onboard lidar and the cloud map it constructs, combined with ultrasonic radar and vehicle-mounted motor encoders and other ranging devices, to determine the position of the stop or return station ahead, and then controls the walking motor to decelerate, reducing the speed of the photovoltaic sweeper and making it move forward slowly. When it reaches the stop or return station, the walking wheels of the photovoltaic sweeper will not wear, avoiding violent collisions between the machine body and the stop or return station, thus improving the overall service life and stability of the equipment.
[0059] 4. The photovoltaic sweeper operation control method and device provided by the present invention adopts point cloud map acquisition by laser radar combined with ultrasonic radar and other sensors to realize point cloud map construction and positioning of photovoltaic sweeper in complex environment, with strong anti-interference ability. Attached Figure Description
[0060] The preferred embodiments will now be described in a clear and easy-to-understand manner, with reference to the accompanying drawings, to further explain the above-mentioned characteristics, technical features, advantages, and implementation methods of a photovoltaic cleaning machine operation control method and device.
[0061] Figure 1 This is a schematic flowchart of an embodiment of the operation control method for a photovoltaic cleaning machine according to the present invention;
[0062] Figure 2 This is a schematic diagram of a steep-slope bridge in another embodiment of the operation control method for a photovoltaic cleaning machine according to the present invention;
[0063] Figure 3 This is a schematic diagram of a large misaligned cable tray in another embodiment of the operation control method for a photovoltaic cleaning machine according to the present invention;
[0064] Figure 4 This is a block diagram of an embodiment of the operation control device for a photovoltaic cleaning machine according to the present invention;
[0065] Figure 5 This is a schematic diagram illustrating the calculation of the bridge deck slope in one embodiment of the operation control device for a photovoltaic cleaning machine according to the present invention.
[0066] Figure label:
[0067] a1--First photovoltaic module; a2--Second photovoltaic module; b--Cable tray; 10--Point cloud map construction module; 20--Slope information acquisition module; 30--Distance measurement module; 40--Speed adjustment module; c1--First ultrasonic radar; c2--Second ultrasonic radar. Detailed Implementation
[0068] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.
[0069] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or sets.
[0070] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.
[0071] Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.
[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings and other implementation methods can be obtained based on these drawings without any creative effort.
[0073] In one embodiment, the present invention provides an operation control method for a photovoltaic cleaning machine, referring to... Figure 1 The steps include:
[0074] S100 uses a lidar installed on a photovoltaic sweeper to construct a point cloud map of the cleaning environment;
[0075] The S200 identifies cable trays ahead based on a point cloud map of the cleaning environment and obtains the slope information of the cable trays.
[0076] When the S300 detects a cable tray, it measures the current distance to the cable tray using a distance sensor installed on the photovoltaic sweeper.
[0077] The S400 adjusts the speed of the photovoltaic sweeper based on the current distance to the cable tray and the slope of the cable tray.
[0078] Specifically, this embodiment uses point cloud mapping to achieve automatic navigation of the photovoltaic sweeper, and can adjust the sweeper's speed near the cable tray based on the type and slope of the cable tray. The photovoltaic sweeper is equipped with a LiDAR, which can pre-build a point cloud map of the sweeper's cleaning route. This point cloud map is then used to identify the cable tray ahead of the sweeper's route, allowing for early detection of locations that could prevent the sweeper from operating properly and facilitating preventative measures. Therefore, it is necessary to first obtain the slope information of the cable tray and the distance between the sweeper and the cable tray, thus adjusting the sweeper's speed in advance to allow for proactive measures. The advantage of this approach is that by pre-building a point cloud map, it is easy to identify and obtain the slope information of the cable tray, and then control the sweeper's speed to prevent it from jamming or tipping over due to excessive cable tray inclination angles.
[0079] In one embodiment, based on the above embodiment, step S100 constructs a point cloud map of the cleaning environment using a lidar installed on the photovoltaic cleaning machine; specifically including:
[0080] At least two environmental point clouds are acquired by scanning with at least two lidars installed on the photovoltaic cleaning machine; the two lidars are on the same horizontal or vertical line.
[0081] Extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud;
[0082] Voxel filtering is used to downsample the truncated effective point cloud;
[0083] Set up NDT, calculate the transformation matrix, and perform NDT registration on at least two downsampled environmental point clouds based on the transformation matrix;
[0084] At least two registered environmental point clouds are stitched together to generate a 3D cleaning environment point cloud map.
[0085] Specifically, raw point cloud data often contains a large number of scattered points and isolated points. Therefore, when acquiring point cloud data, some noise will inevitably appear in the data due to the influence of equipment accuracy and other issues. Noise must be filtered out before use. 3D point clouds often contain a large amount of redundant data, making direct processing computationally intensive and time-consuming. Therefore, it is necessary to downsample the effective point cloud. An NDT (Normal Distribution Transform) is set to register and stitch the downsampled environmental point cloud, and then a 3D cleaned environment point cloud map is generated.
[0086] In another embodiment, based on the above embodiment, step S200 identifies the cable tray ahead based on the point cloud map of the cleaning environment and obtains the slope information of the cable tray; specifically including:
[0087] Based on the point cloud map of the cleaned environment, photovoltaic modules and cable trays in the point cloud map of the cleaned environment were identified;
[0088] Point clouds of photovoltaic modules and cable trays were segmented using a semantic segmentation algorithm;
[0089] Project the point clouds of photovoltaic modules and cable trays to obtain a 2D target projection point cloud map;
[0090] Linearity fitting is performed on the point cloud in the 2D target projection point cloud map to obtain linear expressions for photovoltaic modules and cable trays.
[0091] Based on the linear expressions of photovoltaic modules and cable trays, the relative angle between the cable trays and photovoltaic modules is determined.
[0092] Specifically, during the cleaning operation of the photovoltaic sweeper, a point cloud map is acquired using LiDAR. If there is a cable tray in front, the point cloud map can identify the cable tray and the photovoltaic modules. Target identification can be achieved through cluster analysis and feature comparison, or by using deep learning to identify the photovoltaic modules and cable trays in the point cloud map. Because this point cloud is based on 3D space and has complete three-dimensional spatial information, but when calculating the cable tray angle, only the slope angle θ of the cable tray is considered, the angle between the cable tray and the current module can be measured by projecting the 3D point cloud onto a 2D view. The specific method involves using a semantic segmentation algorithm to segment the point cloud of the cable tray and part of the current module, then projecting this part of the point cloud onto the z-axis, thus obtaining a z-axis point cloud map. A least squares method is then used to perform linear fitting on these point cloud maps to obtain a linear expression for the cable tray and the current module, allowing the calculation of the relative angle between the cable tray and the current module—the slope θ.
[0093] After determining the bridge deck slope, consult the table showing the relationship between bridge deck slope and speed values to calculate the required speed for the sweeper to reach the bridge deck at the current angle. Then, adjust the speed of the travel motor via the controller to achieve the required speed upon arrival at the bridge deck. The control method is closed-loop control (the motor's encoder provides real-time feedback on the current speed). This allows for smooth passage through bridge decks with intersecting components.
[0094] In another embodiment, the slope of the cable tray can be measured and calculated using distance sensors on the same vertical line on the photovoltaic sweeper. The slope of the cable tray can be obtained through the following steps:
[0095] The first distance and the second distance to the front cable tray are measured by the first distance sensor and the second distance sensor installed on the same vertical line on the photovoltaic sweeper, respectively;
[0096] Obtain the difference between the first distance and the second distance;
[0097] Based on the difference, and combined with the installation distance between the first and second ranging sensors, the slope of the cable tray is obtained.
[0098] Specifically, the photovoltaic sweeper is equipped with an ultrasonic radar at both the top and bottom, with the two radars aligned vertically. The distance to the slope in front is measured by the two ultrasonic radars, and then the slope is calculated using trigonometric theorems. For example... Figure 5 As shown, the two ultrasonic radars (first ultrasonic radar c1 and second ultrasonic radar c2) of the photovoltaic sweeper measure the distances from the cable tray as L1 and L2 respectively, and the distance between the two ultrasonic radars is d. The slope of the cable tray in front is calculated as: θ=arctan(d / (L1-L2)).
[0099] In another embodiment, based on the above embodiments, step S400 specifically includes:
[0100] Based on the slope information of the cable tray, the target speed of the photovoltaic sweeper when it reaches the cable tray is found from the pre-stored slope-speed correspondence table;
[0101] Based on the current distance from the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper, the corresponding acceleration is obtained;
[0102] The operation of the photovoltaic sweeper is controlled by acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
[0103] For details, please refer to Table 1.
[0104] Cable tray angle (θ) 0° 5° 10° 15° 20° 25° 30° Target velocity (v) <![CDATA[X1]]> <![CDATA[X2]]> <![CDATA[X3]]> <![CDATA[X4]]> <![CDATA[X5]]> <![CDATA[X6]]> <![CDATA[X7]]>
[0105] Table 1 shows the speed table for the photovoltaic sweeper corresponding to different bridge slopes. The corresponding speeds in the table vary depending on the bridge slope. After obtaining the bridge slope, the required uphill speed—the target speed Vs—is obtained based on the actual tested and calibrated slope-speed table. Then, the distance S between the photovoltaic sweeper and the bridge is obtained from the distance sensor. Combining this with the current speed Vc and the target speed Vs, the required acceleration is calculated using the velocity-displacement formula: Vs^2=Vc^2+2aS: a=(Vs^2-Vc^2) / 2S.
[0106] For some bridge structures with steep slopes (reference) Figure 2 When a photovoltaic sweeper operates at normal speed, it cannot climb slopes. Therefore, the steeper the slope, the higher the target speed. To adjust the speed, ultrasonic radar or other ranging devices can be used to determine the distance from the sweeper's current position to the bridge. Combined with the current speed, the acceleration required to reach the target speed for climbing the slope can be calculated. The sweeper then accelerates towards the bridge according to the calculated acceleration, ensuring it reaches the target speed upon arrival, thus enabling it to climb slopes.
[0107] For reference of the connection bridge of large misaligned components Figure 3 The connecting cable trays of these structures often have large lateral angles. If the photovoltaic sweeper were to pass over these trays at normal operating speed, it could easily become stuck and unable to cross. In this situation, laser radar and constructed cloud maps, combined with ultrasonic radar and ranging devices, can be used to obtain information such as the height of the cable tray ahead, thus calculating the misalignment position. The controller on the photovoltaic sweeper then controls the walking motor to slow down, reducing the sweeper's speed and allowing it to move forward slowly, minimizing the risk of getting stuck when crossing the cable trays.
[0108] In another embodiment, based on any of the above embodiments, the step of:
[0109] Based on the point cloud map of the cleaning environment, identify the stop or return station ahead;
[0110] When a stop or return station is detected, the distance between the photovoltaic sweeper and the stop or return station is measured by a distance sensor installed on the photovoltaic sweeper.
[0111] Based on the distance between the photovoltaic sweeper and the docking or return station, and combined with the current speed of the photovoltaic sweeper, the speed of the photovoltaic sweeper is reduced so that the speed when the photovoltaic sweeper arrives at the docking or return station is less than the preset speed.
[0112] Specifically, by utilizing the lidar mounted on the photovoltaic sweeper and its pre-built point cloud map, combined with ultrasonic radar and onboard motor encoders and other ranging devices, the system identifies the location of the upcoming stopping or return station, as well as the distance between the sweeper and different locations. Upon reaching the vicinity of the bridge, the robot's controller decelerates the walking motor, reducing the sweeper's speed and allowing it to move forward slowly. The advantage of this approach is that it prevents wear on the sweeper's wheels when it reaches the stopping or return station, avoiding violent collisions and improving the overall lifespan and stability of the equipment.
[0113] Based on the same technical concept, the present invention also provides an operation control device for a photovoltaic cleaning machine, see reference. Figure 4 In one embodiment, it includes:
[0114] Point cloud map building module 10 is used to build a point cloud map of the cleaning environment using a lidar installed on a photovoltaic cleaning machine;
[0115] The slope information acquisition module 20 is used to identify the cable tray in front based on the point cloud map of the cleaning environment and to acquire the slope information of the cable tray.
[0116] The distance measurement module 30 is used to measure the current distance to the cable tray when the cable tray is detected by a distance sensor installed on the photovoltaic cleaning machine;
[0117] The speed adjustment module 40 is used to adjust the speed of the photovoltaic sweeper based on the current distance from the cable tray and the slope information of the cable tray.
[0118] Specifically, the point cloud map construction module 10 uses the lidar on the laser photovoltaic sweeper to identify the cleaning route of the photovoltaic sweeper and construct a point cloud map. Then, when the photovoltaic sweeper encounters a special bridge, the slope information acquisition module 20 obtains the slope information of the bridge based on the constructed point cloud map. The distance measurement module 30 measures the distance from the photovoltaic sweeper to the bridge. Finally, the speed adjustment module 40 combines the previously measured bridge slope, the distance between the photovoltaic sweeper and the bridge, and the current speed of the photovoltaic sweeper to control the speed of the photovoltaic sweeper, enabling it to autonomously complete functions such as component crossing.
[0119] In another embodiment, based on the above embodiments, the point cloud map construction module specifically includes:
[0120] The environmental point cloud acquisition submodule is used to acquire at least two environmental point clouds by scanning with at least two lidars installed on the photovoltaic cleaning machine; the two lidars are on the same horizontal or vertical line.
[0121] The noise filtering submodule is used to extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud.
[0122] The downsampling submodule is used to downsample the truncated effective point cloud using voxel filtering.
[0123] The point cloud map construction submodule is used to set up NDT, calculate the transformation matrix, and perform NDT registration on at least two downsampled environmental point clouds based on the transformation matrix; the registered at least two environmental point clouds are then stitched together to generate a 3D cleaning environment point cloud map.
[0124] Specifically, the environmental point cloud acquisition submodule is mainly used to acquire point clouds in the environment, which is accomplished using LiDAR. Raw point cloud data often contains a large number of scattered and isolated points. Therefore, due to limitations in equipment accuracy, some noise is inevitably present in the point cloud data during acquisition. A noise filtering submodule is used to filter out this noise. 3D point clouds often contain a large amount of redundant data, making direct processing computationally intensive and time-consuming. Therefore, a downsampling submodule is used to downsample the effective point cloud data. The point cloud map construction submodule sets up NDT (Normal Distribution Transform), registers and stitches the downsampled environmental point cloud, and then generates a 3D cleaned environment point cloud map.
[0125] In another embodiment, based on the above embodiments, the slope information acquisition module specifically includes:
[0126] The identification submodule is used to identify photovoltaic modules and cable trays in the cleaning environment point cloud map based on the cleaning environment point cloud map;
[0127] The algorithm segmentation submodule is used to segment the point clouds of photovoltaic modules and cable trays using semantic segmentation algorithms;
[0128] The projection submodule is used to project the point cloud of photovoltaic modules and cable trays to obtain a 2D target projection point cloud map;
[0129] The fitting submodule is used to linearize and fit the point cloud in the 2D target projection point cloud map to obtain the linear expressions of photovoltaic modules and cable trays.
[0130] The slope acquisition submodule is used to determine the relative angle between the photovoltaic modules and the photovoltaic modules based on the linear expressions of the photovoltaic modules and the cable trays.
[0131] Specifically, the identification submodule uses deep learning to identify photovoltaic modules and cable trays in the environmental point cloud map. It then segments the identified point cloud using a semantic segmentation algorithm. The projection submodule projects the point clouds of the photovoltaic modules and cable trays to generate a 2D projected point cloud map. The fitting submodule performs linear fitting on the information in the projected point cloud map to obtain a linear expression for the cable trays. Analyzing this expression yields the relative angle between the cable trays and the photovoltaic modules, i.e., the slope information of the cable trays.
[0132] In another embodiment, the slope of the cable tray can be measured and calculated using a distance sensor on the same vertical line on the photovoltaic sweeper.
[0133] The distance measurement module is used to measure the first distance and the second distance from the front cable frame respectively by a first distance sensor and a second distance sensor installed on the same vertical line on the photovoltaic sweeper.
[0134] The slope information acquisition module calculates the difference between the first distance and the second distance; then, based on the difference and the installation distance between the first distance sensor and the second distance sensor, it obtains the slope of the cable tray.
[0135] Specifically, the photovoltaic sweeper is equipped with an ultrasonic radar at the top and bottom, both aligned vertically. The distance to the slope in front is measured by the two ultrasonic radars, and then the slope is calculated using trigonometric theorems. For example, refer to... Figure 5 The two ultrasonic radars (first ultrasonic radar c1 and second ultrasonic radar c2) of the photovoltaic sweeper measure the distances from the cable tray as L1 and L2 respectively, and the distance between the two ultrasonic radars is d. The slope of the cable tray in front is calculated as: θ=arctan(d / (L1-L2)).
[0136] In another embodiment, based on the above embodiments, the speed adjustment module specifically includes:
[0137] The speed acquisition submodule is used to look up the target speed of the photovoltaic sweeper when it reaches the bridge based on the slope information of the bridge.
[0138] The acceleration acquisition submodule is used to obtain the corresponding acceleration based on the current distance to the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper;
[0139] The speed adjustment submodule is used to control the operation of the photovoltaic sweeper based on acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
[0140] Specifically, the acceleration acquisition submodule determines the required acceleration based on the slope information of the cable tray and the current distance between the photovoltaic sweeper and the cable tray. Then, the acceleration adjustment submodule adjusts the acceleration of the photovoltaic sweeper to control its running speed, so that the photovoltaic sweeper can climb and cross the slope normally when it reaches the vicinity of the cable tray.
[0141] In another embodiment, based on any of the above embodiments, it further includes:
[0142] The docking module is used to identify the docking station or return station ahead based on the point cloud map of the cleaning environment.
[0143] When a stop or return station is detected, the distance between the photovoltaic sweeper and the stop or return station is measured by a distance sensor installed on the photovoltaic sweeper.
[0144] Based on the distance between the photovoltaic sweeper and the docking or return station, and combined with the current speed of the photovoltaic sweeper, the speed of the photovoltaic sweeper is reduced so that the speed when the photovoltaic sweeper arrives at the docking or return station is less than the preset speed.
[0145] Specifically, by utilizing onboard lidar and its constructed cloud map, combined with ultrasonic radar and vehicle-mounted motor encoders and other ranging devices, the location of the upcoming docking or return station, as well as the distance between different parts and the photovoltaic sweeper, can be determined. Upon reaching the vicinity of the bridge, the robot's onboard controller decelerates the walking motor, thus reducing the photovoltaic sweeper's speed and allowing it to move forward slowly. This prevents wear on the sweeper's wheels upon reaching the docking or return station, avoids violent collisions between the machine and the station, and improves the overall lifespan and stability of the equipment.
[0146] These can be implemented using computer-executable program code, thus allowing them to be stored in a storage device for execution by a computing device, or fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Therefore, this invention is not limited to any particular hardware and software combination.
[0147] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.
[0148] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0149] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0150] Furthermore, the functional units in the various embodiments of this application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The integrated unit described above can be implemented in hardware or as a software functional unit.
[0151] It should be noted that the above embodiments can be freely combined as needed. The above description is only a preferred embodiment of the present invention. It should be pointed out that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method of operating control of a photovoltaic cleaning machine, characterized by, Including the following steps: A point cloud map of the cleaning environment is constructed using a lidar installed on a photovoltaic cleaning machine; Based on the cleaning environment point cloud map, the system identifies the cable tray ahead and obtains the slope information of the cable tray, including: identifying the photovoltaic modules and cable tray in the cleaning environment point cloud map; segmenting the point clouds of the photovoltaic modules and cable tray using a semantic segmentation algorithm; projecting the point clouds of the photovoltaic modules and cable tray to obtain a 2D target projection point cloud map; linearizing and fitting the point clouds in the 2D target projection point cloud map to obtain a linear expression for the photovoltaic modules and cable tray; and determining the relative angle between the cable tray and the photovoltaic modules based on the linear expression for the photovoltaic modules and cable tray. When the cable tray is detected, the current distance to the cable tray is measured by a distance sensor installed on the photovoltaic cleaning machine; The process of obtaining the slope information further includes: measuring a first distance and a second distance from the cable tray in front using a first distance sensor and a second distance sensor installed on the same vertical line on the photovoltaic sweeper; obtaining the difference between the first distance and the second distance; and obtaining the slope of the cable tray based on the difference and the installation distance between the first distance sensor and the second distance sensor. The speed of the photovoltaic sweeper is adjusted based on the current distance to the cable tray and the slope information of the cable tray.
2. The method of claim 1, wherein, The construction of a point cloud map of the cleaning environment using a lidar installed on a photovoltaic cleaning machine specifically includes: At least two environmental point clouds are acquired by scanning with at least two lidar sensors installed on the photovoltaic cleaning machine; the two lidar sensors are on the same horizontal or vertical line. Extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud; Voxel filtering is used to downsample the truncated effective point cloud; Set up NDT, calculate the transformation matrix, and perform NDT registration on the downsampled at least two environmental point clouds based on the transformation matrix; The registered at least two environmental point clouds are stitched together to generate a 3D cleaning environment point cloud map.
3. The method of claim 1, wherein the method further comprises: The method of adjusting the speed of the photovoltaic cleaning machine based on the current distance to the cable tray and the slope information of the cable tray specifically includes: Based on the slope information of the cable tray, the target speed of the photovoltaic sweeper when it reaches the cable tray is found from the pre-stored slope-speed correspondence table; Based on the current distance to the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper, the corresponding acceleration is obtained; The operation of the photovoltaic sweeper is controlled according to the acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
4. The method of claim 1-3, wherein, Also includes: Based on the point cloud map of the cleaning environment, identify the stop or return station ahead; When the stop or return station is detected, the distance between the photovoltaic sweeper and the stop or return station is measured by a distance sensor installed on the photovoltaic sweeper. Based on the distance between the photovoltaic sweeper and the docking station or return station, and combined with the current speed of the photovoltaic sweeper, the speed of the photovoltaic sweeper is reduced so that the speed at which the photovoltaic sweeper arrives at the docking station or return station is less than a preset speed.
5. An operation control device of a photovoltaic cleaning machine, characterized by, include: The point cloud map building module is used to build a point cloud map of the cleaning environment using a lidar installed on a photovoltaic sweeper. The slope information acquisition module is used to identify the cable tray in front based on the point cloud map of the cleaning environment and to acquire the slope information of the cable tray; The slope information acquisition module specifically includes: an identification submodule, used to identify photovoltaic modules and cable trays in the cleaning environment point cloud map based on the cleaning environment point cloud map; an algorithm segmentation submodule, used to segment the point clouds of the photovoltaic modules and cable trays using a semantic segmentation algorithm; a projection submodule, used to project the point clouds of the photovoltaic modules and cable trays to obtain a 2D target projection point cloud map; a fitting submodule, used to perform linear fitting on the point clouds in the 2D target projection point cloud map to obtain a linear expression for the photovoltaic modules and cable trays; and a slope acquisition submodule, used to determine the relative angle between the cable trays and the photovoltaic modules based on the linear expressions of the photovoltaic modules and cable trays. The distance measurement module is used to measure the current distance to the cable tray when the cable tray is detected, using a distance sensor installed on the photovoltaic sweeper. Obtaining the slope information further includes: measuring a first distance and a second distance to the cable tray using a first distance sensor and a second distance sensor installed on the photovoltaic sweeper on the same vertical line; obtaining the difference between the first distance and the second distance; and obtaining the slope of the cable tray based on the difference and the installation distance between the first and second distance sensors. The speed adjustment module is used to adjust the speed of the photovoltaic sweeper based on the current distance to the cable tray and the slope information of the cable tray.
6. The operation control device of a photovoltaic cleaning machine according to claim 5, wherein The point cloud map construction module specifically includes: An environmental point cloud acquisition submodule is used to acquire at least two environmental point clouds by scanning with at least two lidars installed on the photovoltaic cleaning machine; the two lidars are on the same horizontal or vertical line. The noise filtering submodule is used to extract the effective range of each environmental point cloud and filter out noise in each environmental point cloud. The downsampling submodule is used to downsample the truncated effective point cloud using voxel filtering. The point cloud map construction submodule is used to set NDT, calculate the transformation matrix, and perform NDT registration on the downsampled at least two environmental point clouds based on the transformation matrix; the registered at least two environmental point clouds are then stitched together to generate a 3D cleaning environment point cloud map.
7. The operation control device for a photovoltaic cleaning machine according to claim 6, characterized in that, The speed adjustment module specifically includes: The speed acquisition submodule is used to look up the target speed of the photovoltaic sweeper when it reaches the bridge from a pre-stored slope-speed correspondence table based on the slope information of the bridge. An acceleration acquisition submodule is used to obtain the corresponding acceleration based on the current distance to the cable tray and the target speed, combined with the current speed of the photovoltaic sweeper; The speed adjustment submodule is used to control the operation of the photovoltaic sweeper according to the acceleration so that the photovoltaic sweeper reaches the target speed when it reaches the bridge.
Citation Information
Patent Citations
Photovoltaic cell panel cleaning device as well as slope-crossing control method, control system and array thereof
CN108607823A
Floor sweeping robot climbing control method and device, storage medium and floor sweeping robot
CN111419116A
Systems and methods for pre-hill cruise speed adjustment
US20170129492A1
Self-Moving Device Capable of Automatically Recognizing an Object in Front, and Recognition Method Thereof
US20190227574A1
