Task indication and execution method for photovoltaic cleaning robot and related equipment

Abstract

The present disclosure provides a task indication and execution method of a photovoltaic cleaning robot and related equipment; in the method, the photovoltaic cleaning robot obtains image inference data inferred from images shot by a UAV during execution of a cleaning task, determines whether the cleaning task needs to be paused, and after waiting for a passage identification result to be obtained, determines whether to continue or stop executing the cleaning task, so as to accurately determine whether to continue executing the cleaning task according to whether the passage influencing object can pass, and make the passage identification unnecessary to be observed by a person, and the labor cost of photovoltaic cleaning is reduced.

Description

Technical Field

[0001] This disclosure relates to the field of automation control technology, and in particular to a task instruction and execution method and related equipment for a photovoltaic cleaning robot. Background Technology

[0002] Because photovoltaic (PV) modules are often placed outdoors, dust and other contaminants easily accumulate on their surfaces, leading to a decrease in their power generation efficiency. Therefore, regular cleaning of PV modules is a necessary measure to ensure stable power generation, and cleaning robots are an efficient tool to replace manual cleaning.

[0003] When a cleaning robot moves and cleans on a photovoltaic array, it may encounter objects that obstruct its passage, such as weeds or cable trays. It is necessary to identify the passability of these objects; otherwise, forcing its way through may cause the cleaning robot to get stuck or even become detached.

[0004] Existing technologies typically rely on human observation to determine the accessibility of photovoltaic cleaning robots on photovoltaic arrays, resulting in excessively high labor costs. Summary of the Invention

[0005] In view of this, the purpose of this disclosure is to provide a task instruction and execution method and related equipment for a photovoltaic cleaning robot, so as to reduce the human cost of accessibility identification for photovoltaic cleaning robots.

[0006] In a first aspect, embodiments of this disclosure provide a task execution method for a photovoltaic cleaning robot, applied to a photovoltaic cleaning robot, the method comprising: responding to a cleaning command, starting to execute a cleaning task for a photovoltaic array; the photovoltaic array comprising at least one row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device; during the execution of the cleaning task, acquiring target image inference data; the target image inference data being inferred from a target image of the photovoltaic array taken by a drone at its location; the target image inference data including an image inference result and shooting coordinate information of the shooting location of the target image; the image inference result being used to indicate whether there is a passage obstacle affecting the photovoltaic array at the shooting location; the passage obstacle including the connecting device; if there is a passage obstacle affecting the photovoltaic array at the shooting location, pausing the execution of the cleaning task and waiting for the passageability identification result of the passage obstacle; if the passageability identification result indicates that the passage obstacle is passable, continuing to execute the cleaning task; if the passageability identification result indicates that the passage obstacle is not passable, stopping the execution of the cleaning task.

[0007] Secondly, this disclosure provides a task instruction method for a photovoltaic cleaning robot, applied to a drone equipped with a camera. The method includes: responding to an inspection command, and according to the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot indicated by the inspection command, flying above a first photovoltaic cleaning robot that begins its cleaning task, to begin performing an inspection task for the first photovoltaic cleaning robot; wherein the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and each row of photovoltaic modules carries at most one photovoltaic cleaning robot; flying above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and capturing images through the camera to obtain the target image. The system obtains a target image and the shooting coordinates of the drone's location when the target image is captured; it sends the target image and the shooting coordinates to a target processing device to obtain flight instructions generated based on the target image and the shooting coordinates; the target processing device is used to instruct the photovoltaic cleaning robot or the cloud platform; according to the flight instructions, the drone starts flying along a preset circular route at its current location, continues flying along the movement direction of the first photovoltaic cleaning robot, or flies above the next photovoltaic cleaning robot to begin performing a cleaning task, so as to begin performing an inspection task for the second photovoltaic cleaning robot; when all the photovoltaic cleaning robots stop performing cleaning tasks, all the inspection tasks are stopped.

[0008] Thirdly, this disclosure provides a task instruction method for a photovoltaic cleaning robot, applied to a cloud platform. The method includes: determining the starting coordinates and cleaning sequence of the starting position of at least one photovoltaic cleaning robot in a photovoltaic array; wherein the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; each row of photovoltaic modules carries at most one photovoltaic cleaning robot; according to the starting coordinates and the cleaning sequence, a cleaning instruction is sent to the first photovoltaic cleaning robot that first performs the cleaning task, and an inspection instruction is sent to a drone, wherein the cleaning instruction is used to instruct the first photovoltaic cleaning robot to start performing the cleaning task, and the inspection instruction is used to instruct the drone to start performing an inspection task for the first photovoltaic cleaning robot; Based on the target image of the photovoltaic array captured by the UAV at its location, and the shooting coordinates of the target image's shooting location, target image inference data is generated. The target image inference result is used to indicate whether there are any obstacles affecting the passage of the photovoltaic array at the shooting location. Based on the target image inference data, an execution command is sent to the first photovoltaic cleaning robot, and a flight command is sent to the UAV. The execution command is used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task. The flight command is used to instruct the UAV to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to begin performing a cleaning task, so as to begin performing an inspection task for the second photovoltaic cleaning robot.

[0009] Fourthly, this disclosure provides a task execution method for a photovoltaic cleaning robot, applied to a task execution system for a photovoltaic cleaning robot. The system includes: a drone and at least one photovoltaic cleaning robot; the drone is equipped with a camera; the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, with at most one photovoltaic cleaning robot mounted on each row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device; the method includes: a first photovoltaic cleaning robot responding to a cleaning command to begin executing a cleaning task for the photovoltaic array, the first photovoltaic cleaning robot being the first to execute the cleaning task; the drone responding to an inspection command to begin executing an inspection task for the first photovoltaic cleaning robot; the drone flying above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and capturing images through the camera to obtain a target image, and the shooting coordinate information of the drone's position when capturing the target image; the first photovoltaic cleaning robot performing the cleaning... During the task, target image inference data is acquired; the target image inference data is obtained based on the target image; the target image inference data includes image inference results and shooting coordinate information; the image inference results are used to indicate whether there are obstacles affecting the access of the photovoltaic array at the shooting location; the obstacles affecting the access include the connecting device; if there are obstacles affecting the access of the photovoltaic array at the shooting location, the first photovoltaic cleaning robot suspends the cleaning task and waits for the accessibility recognition result of the obstacles affecting the access; the drone starts flying along a preset circumferential route at its location and takes circumferential observation images; the first photovoltaic cleaning robot acquires the accessibility recognition result generated based on the circumferential observation images, and continues or stops the cleaning task based on the accessibility recognition result; the drone continues the inspection task or starts the inspection task for the second photovoltaic cleaning robot based on the accessibility recognition result; when all photovoltaic cleaning robots stop the cleaning task, the drone stops the inspection task.

[0010] Fifthly, this disclosure provides a task execution device for a photovoltaic cleaning robot, applied to a photovoltaic cleaning robot. The device includes: a first response module, configured to respond to a cleaning command and begin executing a cleaning task for a photovoltaic array; the photovoltaic array includes at least one row of photovoltaic modules; the passage gap between each row of photovoltaic modules is connected by a connecting device; a first acquisition module, configured to acquire target image inference data during the execution of the cleaning task; the target image inference data is obtained by inferring from a target image of the photovoltaic array taken by a drone at its location; the target image inference data includes an image inference result and the target image inference result. The system includes: a shooting coordinate information of the shooting location of the target image; the image inference result used to indicate whether there is a traffic obstruction at the photovoltaic array at the shooting location; the traffic obstruction includes the connecting device; a first waiting module used to pause the cleaning task and wait for the accessibility identification result of the traffic obstruction if there is a traffic obstruction at the photovoltaic array at the shooting location; a first execution module used to continue the cleaning task if the accessibility identification result indicates that the traffic obstruction is passable; and a first stop module used to stop the cleaning task if the accessibility identification result indicates that the traffic obstruction is not passable.

[0011] Sixthly, this disclosure provides a task instruction device for a photovoltaic cleaning robot, applied to a drone. The drone is equipped with a camera. The device includes: a second response module, configured to respond to an inspection command and, based on the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot indicated by the inspection command, fly above a first photovoltaic cleaning robot that has begun its cleaning task, to begin performing an inspection task for the first photovoltaic cleaning robot; wherein the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and at most one photovoltaic cleaning robot is mounted on each row of photovoltaic modules; and a first imaging module, configured to fly above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot and capture images through the camera to obtain a target image. The system includes: a first acquisition module, which receives the target image and the shooting coordinates of the drone when the target image is captured; a second acquisition module, which sends the target image and the shooting coordinates to a target processing device to obtain flight instructions generated based on the target image and the shooting coordinates; the target processing device instructs the photovoltaic cleaning robot or a cloud platform; a third response module, which allows the drone to start flying along a preset circular route, continue flying along the direction of movement of the first photovoltaic cleaning robot, or fly above the next photovoltaic cleaning robot to begin performing a cleaning task, in accordance with the flight instructions, so as to begin performing an inspection task for the second photovoltaic cleaning robot; and a second stop module, which stops performing all inspection tasks when all photovoltaic cleaning robots stop performing cleaning tasks.

[0012] In a seventh aspect, this disclosure provides a task instruction device for a photovoltaic cleaning robot, applied to a cloud platform. The device includes: a first determining module, configured to determine the starting coordinate information and cleaning sequence of the starting position of at least one photovoltaic cleaning robot in a photovoltaic array; wherein the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; and each row of photovoltaic modules can be equipped with at most one photovoltaic cleaning robot; a first sending module, configured to send a cleaning instruction to the first photovoltaic cleaning robot that first performs a cleaning task, and to send an inspection instruction to a drone, based on the starting coordinate information and the cleaning sequence; the cleaning instruction instructs the first photovoltaic cleaning robot to begin performing the cleaning task, and the inspection instruction instructs the drone to begin performing an inspection task for the first photovoltaic cleaning robot; the first... A generation module is used to generate target image inference data based on the target image of the photovoltaic array taken by the UAV at its location and the shooting coordinate information of the shooting location of the target image; the target image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location; a second sending module is used to send an execution command to the first photovoltaic cleaning robot and a flight command to the UAV based on the target image inference data; the execution command is used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task; the flight command is used to instruct the UAV to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to start performing the cleaning task, so as to start performing an inspection task for the second photovoltaic cleaning robot.

[0013] Eighthly, this disclosure provides an electronic device including a processor and a memory. The memory stores machine-executable instructions that can be executed by the processor. The processor executes the machine-executable instructions to implement the above-described task execution method or task instruction method of the photovoltaic cleaning robot.

[0014] Ninthly, embodiments of this disclosure provide a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are invoked and executed by a processor, the computer-executable instructions cause the processor to implement the above-described task execution method or task instruction method of the photovoltaic cleaning robot.

[0015] The embodiments disclosed herein bring the following beneficial effects:

[0016] The aforementioned photovoltaic cleaning robot's task instructions, execution methods, and related equipment allow the robot to acquire image reasoning data derived from images captured by a drone during the cleaning process. This data determines whether the cleaning task needs to be paused, and after obtaining the accessibility recognition result, it decides whether to continue or stop the cleaning task. This accurately determines whether to continue the cleaning task based on whether the accessibility is obstructed, eliminating the need for human observation for accessibility recognition and reducing the labor costs of photovoltaic cleaning.

[0017] Other features and advantages of this disclosure will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the disclosure. The objects and other advantages of this disclosure are realized and obtained through the structures particularly pointed out in the description, claims and drawings.

[0018] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the specific embodiments of this disclosure or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart of one embodiment of the task execution method of the photovoltaic cleaning robot in this disclosure;

[0021] Figure 2 This is a schematic diagram of the task execution method of the photovoltaic cleaning robot in an embodiment of this disclosure;

[0022] Figure 3 This is another schematic diagram of the task execution method of the photovoltaic cleaning robot in the embodiments of this disclosure;

[0023] Figure 4 This is another schematic diagram of the task execution method of the photovoltaic cleaning robot in the embodiments of this disclosure;

[0024] Figure 5 This is a flowchart of one embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure;

[0025] Figure 6 This is a schematic diagram of the task execution method of the photovoltaic cleaning robot in an embodiment of this disclosure;

[0026] Figure 7This is a flowchart of another embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure;

[0027] Figure 8 This is another schematic diagram of the task execution method of the photovoltaic cleaning robot in the embodiments of this disclosure;

[0028] Figure 9 This is another schematic diagram of the task execution method of the photovoltaic cleaning robot in the embodiments of this disclosure;

[0029] Figure 10 This is a flowchart of another embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure;

[0030] Figure 11 This is a schematic diagram of a task execution device for a photovoltaic cleaning robot provided in an embodiment of the present disclosure;

[0031] Figure 12 A schematic diagram of a task instruction device for a photovoltaic cleaning robot provided in an embodiment of this disclosure;

[0032] Figure 13 Another schematic diagram of a task instruction device for a photovoltaic cleaning robot provided in this disclosure embodiment;

[0033] Figure 14 This is a schematic diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0035] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in this disclosure, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] For ease of understanding, the specific process of the embodiments of this disclosure is described below. Please refer to [link / reference]. Figure 1 This disclosure includes one embodiment of a task execution method for a photovoltaic cleaning robot, applied to a photovoltaic cleaning robot, and the embodiment includes:

[0037] Step S101: In response to the cleaning command, begin the cleaning task for the photovoltaic array; the photovoltaic array contains at least one row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device;

[0038] It should be noted that photovoltaic (PV) modules, as the smallest power generation units, are combined in a specific layout and connection method to form a PV array, which can convert solar energy into electrical energy. In addition to containing at least one row of PV modules, a PV array also includes connecting devices for bridging the gaps between the PV modules. These connecting devices allow a PV cleaning robot to move along a row of PV modules to complete the cleaning task. As an example, and not a limitation, such as... Figure 2 The diagram shows a photovoltaic array comprising a row of photovoltaic modules. This row of photovoltaic modules includes photovoltaic modules 202 and connecting devices 201 that connect the gaps between the photovoltaic modules. A row of photovoltaic modules can contain any number of photovoltaic modules and can be divided into at least two sections 202, each connected by the connecting devices. A photovoltaic cleaning robot 203 moves and cleans on this row of photovoltaic modules. The connecting devices are passageways for the photovoltaic cleaning robot 203, facilitating its movement from one section to another for cleaning. It is understood that deformation of the connecting devices may affect the passage of the photovoltaic cleaning robot. Therefore, the connecting devices are considered passage obstructions. This embodiment of the present disclosure can make a decision on whether to continue passage based on the passageability of the obstructions, thereby avoiding abnormalities caused by the photovoltaic cleaning robot forcibly passing through. In this embodiment, the connecting devices can be any structural device for the passage of the photovoltaic cleaning robot, such as rigid cable trays, flexible cable trays, or clips, etc., and are not specifically limited here.

[0039] In one implementation, the cleaning command can be triggered on the photovoltaic cleaning robot, such as through physical or virtual buttons on the robot; it can also be triggered through a remote control platform, such as a cloud platform or server, etc., without limitation here. In one implementation, the cleaning command is used to trigger the photovoltaic cleaning robot to perform a cleaning task for the photovoltaic array. The cleaning task can be for one row of photovoltaic modules in the photovoltaic array, or for at least two rows of photovoltaic modules in the photovoltaic array. If it is for at least two rows of photovoltaic modules, after the photovoltaic cleaning robot completes the cleaning task for one row of photovoltaic modules, it can move from the track set between the two rows of photovoltaic modules to the other row of photovoltaic modules to continue the cleaning task. Alternatively, a cleaning shuttle can move the photovoltaic cleaning robot to the other row of photovoltaic modules to continue the cleaning task, without limitation here.

[0040] In one implementation, the photovoltaic cleaning robot is connected to a remote control platform (such as a cloud platform). The remote control platform can send cleaning instructions and assign cleaning tasks to the photovoltaic cleaning robot. In this step, the photovoltaic cleaning robot responds to the cleaning instructions sent by the remote control platform and begins to execute the cleaning task for the photovoltaic array.

[0041] Step S102: During the cleaning task, target image inference data is acquired. The target image inference data is obtained by inferring from the target image of the photovoltaic array taken by the drone at its location. The target image inference data includes the image inference result and the shooting coordinate information of the shooting location of the target image. The image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location. Obstacles affecting the passage include connecting devices.

[0042] In this embodiment, during the cleaning task performed by the photovoltaic cleaning robot, the latest target image inference data can be acquired at a preset frequency, for example, every 1 second or every 10 seconds. The latest target image inference data refers to data inferred from the latest target image captured by the drone. It is understood that the drone can assist the photovoltaic cleaning robot in performing the cleaning task by providing macroscopic image data of the photovoltaic array, thus helping the robot to better complete the cleaning task. It should be noted that the drone is equipped with a camera. The target image can be an image captured by the drone while flying before the photovoltaic cleaning robot performs the cleaning task, or an image captured by the drone while flying during the cleaning task. The former reduces the latency of acquiring the target image inference data, while the latter improves the real-time effectiveness of the target image inference data; the specific method is not limited here.

[0043] In this embodiment, the target image inference data can be generated automatically by the photovoltaic cleaning robot or obtained directly from the remote control platform. If obtained from the remote control platform, the method by which the remote control platform generates the target image inference data is the same as that by the photovoltaic cleaning robot. When obtaining from the remote control platform, the target image captured by the drone is sent to the remote control platform, which then generates the target image inference data and sends it to the photovoltaic cleaning robot. The target image inference data includes image inference results and shooting coordinate information. The shooting coordinate information refers to the position coordinates of the drone when it captured the target image. In one embodiment, the shooting coordinate information can be used to determine whether the photovoltaic cleaning robot is within the waiting range of any connecting device, thereby quickly determining whether the photovoltaic cleaning robot needs to pause its cleaning task and wait for the accessibility recognition result for the connecting device.

[0044] It is understandable that obstacles affecting passage include not only connecting devices, but also objects that may affect the passage of the photovoltaic cleaning robot, such as weeds, photovoltaic modules that have protruded or shifted. By identifying the passability of obstacles affecting passage, it can be determined whether the photovoltaic cleaning robot should continue to move, thus avoiding damage, jamming, or even detachment from the photovoltaic array caused by the photovoltaic cleaning robot forcibly passing through.

[0045] In one embodiment, when the target image inference data is generated autonomously by the photovoltaic cleaning robot, the photovoltaic cleaning robot communicates with the drone. Step S102 includes: Step S1021, during the cleaning task, generating target image inference data based on the target image of the photovoltaic array captured by the drone at its location and the shooting coordinates of the target image's shooting position. The drone can communicate bidirectionally with the photovoltaic cleaning robot. After the photovoltaic cleaning robot starts performing the cleaning task, it sends an inspection command to the drone. The drone responds to the inspection command and begins to perform an inspection task for the photovoltaic cleaning robot. During the inspection task, the drone flies above the photovoltaic cleaning robot and captures target images at a preset frequency, recording the shooting coordinates of the target image's shooting position. The photovoltaic cleaning robot acquires the target image and shooting coordinates to generate target image inference data.

[0046] In one embodiment, step S1021 includes: during the cleaning task, constructing a spatial image model of the target image of the photovoltaic array captured by the drone at its location to obtain a current spatial image model; based on the current spatial image model and the shooting coordinates of the target image's shooting location, determining whether there are any obstacles affecting passage at the photovoltaic array at the shooting location, and obtaining target image inference data. In this embodiment, the spatial image model can accurately identify obstacles affecting passage, thereby improving the accuracy of the photovoltaic cleaning robot's accessibility identification.

[0047] In one embodiment, step S1021 includes: acquiring a target image of the photovoltaic array taken by the drone at its location during the cleaning task; the drone flying in the direction of movement of the photovoltaic cleaning robot and taking images at regular intervals; determining whether there are any obstacles affecting the photovoltaic array at the shooting location based on a preset reference photovoltaic array spatial image model, the target image, and the shooting coordinates of the target image's shooting location, and obtaining target image inference data; the reference photovoltaic array spatial image model is constructed based on a reference photovoltaic array that is passable by obstacles. In this embodiment, the photovoltaic cleaning robot compares the preset reference photovoltaic array spatial image model with the target image to determine whether there are any obstacles affecting the photovoltaic array at the shooting location. The reference photovoltaic array spatial image model is constructed based on a reference photovoltaic array that is passable by obstacles. That is, the reference photovoltaic array spatial image model is constructed for a reference photovoltaic array, which includes connecting devices, all of which are passable, and the reference photovoltaic array does not contain any foreign objects other than the connecting devices. Therefore, by comparing the preset benchmark photovoltaic array spatial image model with the target image, it is possible to determine whether there are connecting devices or foreign objects other than connecting devices at the photovoltaic array at the shooting location, that is, whether there are obstacles affecting passage at the photovoltaic array at the shooting location, thereby obtaining target image inference data.

[0048] In one embodiment, the target image inference data includes first image inference data and second image inference data. Based on a preset reference photovoltaic array spatial image model, the target image, and the shooting coordinates of the target image's shooting location, determining whether there are any obstacles affecting traffic at the photovoltaic array at the shooting location to obtain the target image inference data includes: determining whether there are any connecting devices at the photovoltaic array at the shooting location based on the shooting coordinates of the target image's shooting location to obtain the first image inference data; and determining whether there are any obstacles affecting traffic at the photovoltaic array at the shooting location other than connecting devices based on the preset reference photovoltaic array spatial image model and the target image to obtain the second image inference data.

[0049] It is understandable that, since the connecting device is usually located at a fixed position on the photovoltaic array, it can be identified based on the shooting coordinate information. Specifically, the shooting coordinate information is compared with preset connecting device coordinate information to determine whether a connecting device exists on the photovoltaic array at the shooting location, thus obtaining first image inference data. This first image inference data indicates whether a connecting device exists on the photovoltaic array at the shooting location. In one embodiment, if the shooting coordinate information falls within a preset range of the preset connecting device coordinate information, it can be determined that a connecting device exists on the photovoltaic array at the shooting location; otherwise, it can be determined that no connecting device exists. For obstacles affecting traffic other than connecting devices, it is necessary to combine a reference photovoltaic array spatial image model and the target image for identification; specific details are not limited here.

[0050] Step S103: If there are obstacles affecting passage at the photovoltaic array location, pause the cleaning task and wait for the passage recognition results of the obstacles.

[0051] In this embodiment, if there are obstacles affecting passage at the photovoltaic array location, it is necessary to further determine whether the obstacles are passable. In this case, the photovoltaic cleaning robot pauses its cleaning task at the shooting location or a preset waiting location, awaiting the passability recognition result for the obstacles. In one embodiment, a waiting location marker can be set at a preset distance from each connecting device. When the photovoltaic cleaning robot reaches the location of the waiting location marker, it pauses its cleaning task and waits for the passability recognition result for the connecting device. The location of the waiting location marker can be determined based on target image inference data. Combined with the distance the photovoltaic cleaning robot has moved, it can be identified whether the photovoltaic cleaning robot has reached the location of the waiting location marker, thus improving the recognition efficiency of the connecting device.

[0052] It is understood that the accessibility recognition result can be generated automatically by the photovoltaic cleaning robot or obtained directly from the remote control platform. If obtained directly from the remote control platform, the method by which the remote control platform generates the accessibility recognition result is the same as that of the photovoltaic cleaning robot. In one embodiment, if the accessibility recognition result is generated automatically by the photovoltaic cleaning robot, step S103 includes: step S1031, if there are obstacles affecting access to the photovoltaic array at the shooting location, the cleaning task is paused, and the drone is allowed to fly around and capture the surrounding observation image according to the preset surrounding route; step S1032, the surrounding observation image captured by the drone is acquired, and a spatial image model is constructed on the surrounding observation image to obtain the observation spatial image model; step S1033, the accessibility recognition result is obtained by judging whether the obstacles affecting access are passable through the observation spatial image model.

[0053] In this embodiment, if there is an obstacle affecting passage at the photovoltaic array location, the photovoltaic cleaning robot suspends its cleaning task and sends a flight command to the drone, instructing the drone to fly along a preset orbital route to capture the surrounding observation images. The surrounding observation images captured by the drone can be sent to a remote control platform or to the photovoltaic cleaning robot. The photovoltaic cleaning robot acquires the surrounding observation images from the remote control platform or the drone, constructs a spatial image model of the surrounding observation images, and determines whether the obstacle is passable based on the constructed observation spatial image model, thus obtaining a passability identification result. In one embodiment, if the obstacle is a foreign object other than the connecting device, the height of the foreign object is determined to be greater than a preset height threshold based on the foreign object spatial image model in the observation spatial image model. If the height of the foreign object is greater than the preset height threshold, it is determined that the foreign object is not passable; otherwise, it is determined that the foreign object is passable.

[0054] In one embodiment, if the obstacle affecting passage is a connecting device, step S1033 includes: if the obstacle affecting passage is a connecting device, then based on the observed spatial image model, determining whether the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than a preset tilt angle difference threshold; if the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than the preset tilt angle difference threshold, then determining that the passageability identification result indicates that the obstacle affecting passage is not passable; if the tilt angle difference between the photovoltaic modules on both sides of the connecting device is less than or equal to the preset tilt angle difference threshold, then determining that the passageability identification result indicates that the obstacle affecting passage is passable. In this embodiment, if the obstacle affecting passage is a connecting device, determining whether the connecting device is passable can be done by judging whether the tilt angle difference between the photovoltaic modules on both sides of the connecting device in the observed spatial image model is greater than the preset tilt angle difference threshold. If the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than the preset tilt angle difference threshold, then the obstacle affecting passage is determined to be not passable; otherwise, the obstacle affecting passage is determined to be passable. In this embodiment, the passability of the connecting device is determined by the difference in tilt angle between the photovoltaic modules on both sides of the connecting device. Compared with the degree of deformation of the connecting device, this method can more accurately and directly reflect the passability of the connecting device. Furthermore, since the passability identification is not performed on the connecting device itself, it is applicable to connecting devices of any structure and those that undergo any form of deformation, thus having a wide range of applications.

[0055] As an example, and not a limitation, such as Figure 3 and Figure 4 The image shown is a partial or complete schematic diagram of a row of photovoltaic modules. Assuming... Figure 3 and Figure 4 The connecting device 201 is a flexible cable tray. Figure 3 The diagram shows a scenario where the flexible cable tray can be used, meaning the tilt angle difference between the photovoltaic modules 202 on both sides of the flexible cable tray is less than or equal to a preset tilt angle difference threshold. Figure 3 As can be seen, the photovoltaic cleaning robot can still pass through even with slight deformation of the flexible cable tray. Figure 4 The diagram shows a situation where the flexible cable tray cannot be used, specifically when the tilt angle difference between the photovoltaic modules 202 on both sides of the flexible cable tray exceeds a preset tilt angle difference threshold. Figure 4 It is evident that the flexible cable tray underwent severe deformation, making it difficult for the photovoltaic cleaning robot to pass through.

[0056] Step S104: If the accessibility identification result indicates that the accessibility-affecting object is passable, then continue to perform the cleaning task;

[0057] It is understandable that even if there are obstacles affecting the passage of the photovoltaic array at the shooting location, these obstacles may still be passable. For example, small weeds, slight protrusions of photovoltaic modules, or slight misalignment of photovoltaic modules on both sides of the connecting device are not enough to cause abnormalities such as damage, jamming, or even detachment from the photovoltaic array to the photovoltaic cleaning robot. Therefore, if the passage recognition result indicates that the obstacle is passable, the photovoltaic cleaning robot will continue to move and perform the cleaning task.

[0058] Step S105: If the accessibility identification result indicates that the accessibility-affecting object is not passable, then stop the cleaning task.

[0059] Understandably, if the accessibility identification result indicates that the obstacle is impassable, the photovoltaic cleaning robot will stop performing its cleaning task and cease moving forward to avoid damage, jamming, or even detachment from the photovoltaic array caused by forced passage. In one embodiment, after stopping its cleaning task, the photovoltaic cleaning robot can remain in place or return to a preset parking position. The preset parking position can be the starting position of the photovoltaic cleaning robot's cleaning task or a preset reversing position; the specific location is not limited here.

[0060] The task execution method of the photovoltaic cleaning robot provided in the above embodiments allows the photovoltaic cleaning robot to acquire image reasoning data inferred from images captured by drones during the cleaning task, determine whether to pause the cleaning task, and wait for the accessibility recognition result before determining whether to continue or stop the cleaning task. This accurately determines whether to continue the cleaning task based on whether the accessibility obstacle can be passed, so that accessibility recognition does not require human observation, thus reducing the labor cost of photovoltaic cleaning.

[0061] Please see Figure 5 One embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure is applied to a drone equipped with a camera. The embodiment includes:

[0062] Step S501: In response to the inspection command, according to the starting coordinate information and cleaning sequence of the starting position of at least one photovoltaic cleaning robot indicated by the inspection command, fly to above the first photovoltaic cleaning robot that starts to perform the cleaning task, so as to start the inspection task for the first photovoltaic cleaning robot; wherein, the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and at most one photovoltaic cleaning robot is mounted on each row of photovoltaic modules.

[0063] In this embodiment, the inspection command can be sent by a remote control platform or by the photovoltaic cleaning robot. One drone can perform inspection tasks for one photovoltaic cleaning robot or for more than one photovoltaic cleaning robot. When performing inspection tasks for more than one photovoltaic cleaning robot, the drone performs the inspection tasks for each photovoltaic cleaning robot one by one according to the cleaning order. Based on the starting position of the photovoltaic cleaning robot, the drone first flies to the first photovoltaic cleaning robot that starts to perform the cleaning task and performs the inspection task for the first photovoltaic cleaning robot.

[0064] Step S502: Fly above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and take pictures through the camera to obtain the target image and the shooting coordinate information of the drone's position when taking the target image;

[0065] In this embodiment, when the drone performs an inspection task for the first photovoltaic cleaning robot, it flies above the first photovoltaic cleaning robot along its direction of movement at a constant speed. During flight, it captures images at a preset frequency to obtain target images and the coordinate information of the image capture positions, which are used to generate target image inference data. In one embodiment, when flying above the first photovoltaic cleaning robot along its direction of movement, the drone flies in tracking mode. Tracking mode refers to the drone's flight mode in which the first photovoltaic cleaning robot remains within the camera's field of view. Further, the tracking mode can be a fixed-distance tracking mode, which means that the drone flies slightly above and in front of the cleaning robot while maintaining a preset horizontal distance in its direction of movement. This is an example, not a limitation, of... Figure 6 The diagram shows a drone 600 flying above a photovoltaic cleaning robot. The drone 600 maintains a preset horizontal distance from the photovoltaic cleaning robot 203. That is, when the photovoltaic cleaning robot 203 moves, the drone 600 also moves with it, so that the drone can take real-time pictures of whether the photovoltaic array in front of the photovoltaic cleaning robot is passable.

[0066] Step S503: Send the target image and shooting coordinate information to the target processing device to obtain flight instructions generated based on the target image and shooting coordinate information; the target processing device is used to instruct the photovoltaic cleaning robot or cloud platform;

[0067] In this embodiment, the target image and shooting coordinate information captured by the drone can be sent to the photovoltaic cleaning robot or to the remote control platform. The photovoltaic cleaning robot or the remote control platform generates target image inference data based on the target image of the photovoltaic array captured by the drone at its location and the shooting coordinate information of the target image's shooting location; and generates flight commands based on the target image inference data and sends them to the drone to instruct the drone's flight mode.

[0068] Step S504: According to the flight instructions, the UAV starts flying along the preset circular route at its current location, continues to fly along the moving direction of the first photovoltaic cleaning robot, or flies to the top of the second photovoltaic cleaning robot that starts to perform the cleaning task, so as to start performing the inspection task for the second photovoltaic cleaning robot.

[0069] In this embodiment, the UAV responds to the flight command and starts flying along a preset circular route from its current location, continues to fly along the movement direction of the first photovoltaic cleaning robot, or flies above the second photovoltaic cleaning robot that begins to perform the cleaning task, so as to start performing the inspection task for the second photovoltaic cleaning robot.

[0070] Step S505: When all photovoltaic cleaning robots stop performing cleaning tasks, all inspection tasks will also stop.

[0071] In this embodiment, when all photovoltaic cleaning robots stop performing cleaning tasks, the drone also stops performing inspection tasks. After the drone stops performing inspection tasks, it can fly back to the preset parking bay. The parking bay can be set inside the body of any photovoltaic cleaning robot or at any location outside the body of the photovoltaic cleaning robot. The specific location is not limited here.

[0072] The task instruction method for photovoltaic cleaning robots provided by the above embodiments allows drones to perform inspection tasks for one or more photovoltaic cleaning robots, resulting in low inspection costs. Furthermore, the macroscopic image data provided to the photovoltaic cleaning robots enables more accurate identification of the accessibility of the photovoltaic cleaning robots on the photovoltaic array.

[0073] Please see Figure 7 Another embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure is applied to a cloud platform, and the embodiment includes:

[0074] Step S701: Determine the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array; wherein, the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; and each row of photovoltaic modules can be equipped with at most one photovoltaic cleaning robot.

[0075] In this embodiment, the cloud platform can instruct the drone to fly along a preset modeling route to obtain modeling images of the photovoltaic array taken by the drone during the flight of the drone along the preset modeling route. Based on the modeling images, the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array can be determined. The cloud platform can also directly input the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array, which is not limited here.

[0076] In one implementation, the cloud platform can determine the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array based on the modeled image. Specifically, step S701 includes: sending a modeling flight command to the drone to instruct the drone to fly along a preset modeling route and take modeling images; constructing a spatial image model of the photovoltaic array to be cleaned based on the modeling image to obtain a target photovoltaic array spatial image model; determining the starting coordinates of each photovoltaic cleaning robot in the photovoltaic array based on the target photovoltaic array spatial image model; and determining the cleaning sequence based on the starting coordinates.

[0077] In this embodiment, the modeling images taken by the UAV following the modeling flight path can cover the entire photovoltaic array. This is an example, not a limitation. Figure 8 The image shown is a schematic diagram of a preset modeling route. Figure 8 In the diagram, the photovoltaic array to be cleaned consists of four rows of photovoltaic modules, with a photovoltaic cleaning robot mounted on each row. The dotted line with arrows in the diagram represents a preset modeling flight path. The UAV 600 flies along this preset modeling flight path, capturing modeling images at a preset frequency as it flies, thus obtaining a modeling image covering the entire photovoltaic array. Based on the modeling images, a spatial image model is constructed, resulting in the spatial image model of the target photovoltaic array. This allows the determination of the starting coordinates of each photovoltaic cleaning robot's starting position and the cleaning sequence. It should be noted that the construction method for the baseline photovoltaic array spatial image model is the same as that for the target photovoltaic array spatial image model; the only difference is the type of photovoltaic array being cleaned. The baseline photovoltaic array spatial image model is for the baseline photovoltaic array, while the target photovoltaic array spatial image model is for the photovoltaic array to be cleaned. Otherwise, the construction methods are identical and will not be elaborated further here.

[0078] Step S702: Based on the starting coordinate information and cleaning sequence, send a cleaning command to the first photovoltaic cleaning robot that first performs the cleaning task, and send an inspection command to the drone. The cleaning command is used to instruct the first photovoltaic cleaning robot to start performing the cleaning task, and the inspection command is used to instruct the drone to start performing the inspection task for the first photovoltaic cleaning robot.

[0079] In this embodiment, after determining the starting coordinate information and cleaning sequence, the cloud platform sends a cleaning command to the first photovoltaic cleaning robot that is the first to perform the cleaning task, instructing the first photovoltaic cleaning robot to start performing the cleaning task; and sends an inspection command to the drone, instructing the drone to fly from the parking bay to the position indicated by the starting coordinate information of the first photovoltaic cleaning robot, and start performing the inspection task for the first photovoltaic cleaning robot.

[0080] Step S703: Based on the target image of the photovoltaic array taken by the UAV at its location and the shooting coordinate information of the shooting location of the target image, generate target image inference data; the target image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location.

[0081] In this embodiment, the target image of the photovoltaic array captured by the drone, along with the shooting coordinates of the target image's location, is sent to the cloud platform. The cloud platform then generates target image inference data based on the target image and the shooting coordinates. It should be noted that the method by which the cloud platform generates target image inference data is the same as that used by the photovoltaic cleaning robot; details will not be elaborated here.

[0082] Step S704: Based on the target image reasoning data, send an execution command to the first photovoltaic cleaning robot and a flight command to the drone; the execution command is used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task; the flight command is used to instruct the drone to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to begin performing the cleaning task, so as to start performing the inspection task for the second photovoltaic cleaning robot.

[0083] In this embodiment, after the cloud platform generates target image inference data, it can determine the execution command to be sent to the first photovoltaic cleaning robot and the flight command to be sent to the drone. The execution command instructs the first photovoltaic cleaning robot to pause, continue, or stop performing its cleaning task, and the flight command instructs the drone to begin flying along a preset circular route from its current location, continue flying along the direction of movement of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to begin performing its cleaning task, thus initiating an inspection task for the second photovoltaic cleaning robot. When all photovoltaic cleaning robots stop performing their cleaning tasks, the cloud platform sends a stop inspection task execution command to the drone, and the drone stops performing all inspection tasks.

[0084] In one embodiment, step S704 includes: step S7041, if there are obstacles affecting the passage of the photovoltaic array at the shooting location, a first execution command is sent to the first photovoltaic cleaning robot, and a first flight command is sent to the drone, instructing the first photovoltaic cleaning robot to suspend the cleaning task, and instructing the drone to start flying along a preset circumferential route at its current location, and to take circumferential observation images; step S7042, if there are no obstacles affecting the passage of the photovoltaic array at the shooting location, a second execution command is sent to the first photovoltaic cleaning robot, and a second flight command is sent to the drone, instructing the first photovoltaic cleaning robot to continue the cleaning task, and instructing the drone to continue flying along the movement direction of the first photovoltaic cleaning robot.

[0085] In this embodiment, the execution commands sent by the cloud platform to the first photovoltaic cleaning robot and the flight commands sent to the drone differ depending on the result indicated by the target image inference data. If the target image inference result indicates that there are obstacles obstructing traffic at the photovoltaic array location, the cloud platform instructs the first photovoltaic cleaning robot to suspend its cleaning task and instructs the drone to begin flying along a preset circling route from its current location, capturing circling observation images. If the target image inference result indicates that there are no obstacles obstructing traffic at the photovoltaic array location, the cloud platform instructs the first photovoltaic cleaning robot to continue its cleaning task and instructs the drone to continue flying along the direction of movement of the first photovoltaic cleaning robot. The cloud platform allows for unified control of the task execution methods of both the photovoltaic cleaning robot and the drone, improving convenience.

[0086] In one implementation, after step S7041, the following steps are included: acquiring a surround observation image captured by the UAV; constructing a spatial image model from the surround observation image to obtain an observation spatial image model; and determining whether a traffic obstacle is passable using the observation spatial image model to obtain a passability identification result. In this implementation, if the target image inference result indicates that a photovoltaic array at the shooting location has a traffic obstacle, the cloud platform instructs the UAV to begin flying along a preset surround route at its location, and after capturing a surround observation image, acquires the surround observation image captured by the UAV, constructs an observation spatial image model of the surround observation image, and then determines whether a traffic obstacle is passable to obtain a passability identification result. This is an example, not a limitation, of... Figure 9 The dotted line with arrows in the middle shows a schematic diagram of a preset circumferential route. In the diagram, the photovoltaic cleaning robot 203 waits at its location, while the drone 600 captures circumferential observation images and sends them to the cloud platform. The cloud platform constructs an image model of the observation space and determines whether the obstacles affecting passage are passable, thus obtaining the passability recognition result.

[0087] In one implementation, the cloud platform can instruct the photovoltaic cleaning robot and the drone to execute different commands based on different accessibility identification results. Specifically, after obtaining the accessibility identification results, the method further includes: if the accessibility identification results indicate that the accessibility obstacle is passable, instructing the first photovoltaic cleaning robot to continue performing the cleaning task and instructing the drone to continue flying along the movement direction of the first photovoltaic cleaning robot; if the accessibility identification results indicate that the accessibility obstacle is not passable, instructing the first photovoltaic cleaning robot to stop the cleaning task and instructing the drone to fly above the second photovoltaic cleaning robot that will start performing the cleaning task next, so as to start performing the inspection task for the second photovoltaic cleaning robot.

[0088] In this embodiment, if the accessibility identification result indicates that the accessibility obstacle is passable, the cloud platform instructs the photovoltaic cleaning robot to continue performing the cleaning task and instructs the drone to continue flying along the movement direction of the first photovoltaic cleaning robot. If the accessibility identification result indicates that the accessibility obstacle is not passable, the first photovoltaic cleaning robot is instructed to stop the cleaning task, and the drone is instructed to stop performing the inspection task for the first photovoltaic cleaning robot. If there are still photovoltaic cleaning robots that have not performed cleaning tasks, the drone flies to the second photovoltaic cleaning robot that has started performing cleaning tasks to start performing the inspection task for the second photovoltaic cleaning robot. If there are no photovoltaic cleaning robots that have not performed cleaning tasks, that is, all photovoltaic cleaning robots have stopped performing cleaning tasks, the drone also stops performing the inspection task and can fly back to the designated parking bay.

[0089] The task instruction method for photovoltaic cleaning robots provided in the above embodiments allows the cloud platform to uniformly allocate tasks to photovoltaic cleaning robots and drones, process relatively complex image reasoning, generate image reasoning data, and improve the execution and instruction efficiency of cleaning tasks.

[0090] Please see Figure 10 Another embodiment of the task instruction method for the photovoltaic cleaning robot in this disclosure is applied to the task execution system of the photovoltaic cleaning robot. The system includes: a drone and at least one photovoltaic cleaning robot; the drone is equipped with a camera; the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and at most one photovoltaic cleaning robot is mounted on each row of photovoltaic modules; the passage gap between each row of photovoltaic modules is connected by a connecting device; the embodiment includes:

[0091] Step S1001: The first photovoltaic cleaning robot responds to the cleaning command and begins to perform the cleaning task for the photovoltaic array. The first photovoltaic cleaning robot is the first photovoltaic cleaning robot to perform the cleaning task.

[0092] Step S1002: The drone responds to the inspection command and begins to perform the inspection task for the first photovoltaic cleaning robot;

[0093] Step S1003: The drone flies above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and takes pictures through the camera to obtain the target image and the shooting coordinate information of the drone's position when taking the target image.

[0094] Step S1004: During the cleaning task, the first photovoltaic cleaning robot acquires target image reasoning data; the target image reasoning data is obtained by reasoning from the target image; the target image reasoning data includes image reasoning results and shooting coordinate information; the image reasoning results are used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location; obstacles affecting the passage include connecting devices.

[0095] Step S1005: If there are obstacles affecting passage at the photovoltaic array at the shooting location, the first photovoltaic cleaning robot will suspend the cleaning task and wait for the passage recognition result of the obstacles.

[0096] Step S1006: The UAV begins to fly along a preset orbital route from its current location and takes orbital observation images.

[0097] Step S1007: The first photovoltaic cleaning robot obtains the accessibility recognition result generated based on the surrounding observation image, and continues or stops performing the cleaning task based on the accessibility recognition result;

[0098] Step S1008: Based on the accessibility recognition results, the drone continues to perform the inspection task or begins to perform the inspection task for the second photovoltaic cleaning robot.

[0099] Step S1009: When all photovoltaic cleaning robots stop performing cleaning tasks, the drones stop performing all inspection tasks.

[0100] In this embodiment, a drone works in conjunction with at least one photovoltaic (PV) cleaning robot to complete a cleaning task. After receiving a cleaning command, any PV cleaning robot begins cleaning the PV array. The drone, responding to an inspection command from a remote control platform or the PV cleaning robot, begins inspecting the PV cleaning robot, flying above it and capturing images of the target area, recording the coordinates of the captured images. During the cleaning process, the PV cleaning robot can generate its own target image inference data or obtain it from the remote control platform to determine if there are any obstacles obstructing traffic at the PV array location. If obstacles are present, the PV cleaning robot pauses its cleaning task and awaits the accessibility identification results. These results can be obtained from the remote control platform or generated by the PV cleaning robot itself. While the PV cleaning robot waits, the drone flies along a preset orbital path from its current location, capturing circumnavigational images to generate accessibility identification results. The surrounding observation image can be sent to a remote control platform or to the photovoltaic cleaning robot. The target processing device (remote control platform or photovoltaic cleaning robot) receiving the surrounding observation image generates a accessibility recognition result based on the image. The photovoltaic cleaning robot then continues or stops its cleaning task based on the accessibility recognition result. The drone, based on the accessibility recognition result, continues its inspection task or begins an inspection task targeting a second photovoltaic cleaning robot. When all photovoltaic cleaning robots stop cleaning, the drone also stops its inspection task and can return to its docking bay. This implementation allows the photovoltaic cleaning robot and the drone to better cooperate in completing the cleaning task, and the photovoltaic cleaning robot achieves high accessibility recognition accuracy on the photovoltaic array.

[0101] For the corresponding method embodiments described above, see [link to relevant documentation]. Figure 11The diagram shows a task execution device for a photovoltaic cleaning robot, applicable to a photovoltaic cleaning robot. The device includes: a first response module 1100, used to respond to a cleaning command and begin executing a cleaning task targeting a photovoltaic array; the photovoltaic array comprises at least one row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device; and a first acquisition module 1110, used to acquire target image inference data during the execution of the cleaning task; the target image inference data is obtained by inferring from a target image of the photovoltaic array taken by a drone at its location; the target image inference data includes the image inference result and the captured target image. The image includes the shooting coordinates of the location; the image reasoning result is used to indicate whether there is a traffic obstruction at the photovoltaic array at the shooting location; the traffic obstruction includes the connecting device; a first waiting module 1120 is used to pause the cleaning task and wait for the accessibility identification result of the traffic obstruction if there is a traffic obstruction at the photovoltaic array at the shooting location; a first execution module 1130 is used to continue the cleaning task if the accessibility identification result indicates that the traffic obstruction is passable; a first stop module 1140 is used to stop the cleaning task if the accessibility identification result indicates that the traffic obstruction is not passable.

[0102] Optionally, the photovoltaic cleaning robot is communicatively connected to the drone; the first acquisition module 1110 includes a second generation submodule, used to generate target image inference data during the execution of the cleaning task, based on the target image of the photovoltaic array captured by the drone and the shooting coordinate information of the shooting position of the target image.

[0103] Optionally, the second generation submodule includes: an acquisition unit, used to acquire a target image of the photovoltaic array taken by the drone at its location during the execution of the cleaning task; the drone flies in the direction of movement of the photovoltaic cleaning robot and takes images at regular intervals; a judgment unit, used to determine whether there are any obstacles affecting the photovoltaic array at the shooting location based on a preset reference photovoltaic array spatial image model, the target image, and the shooting coordinate information of the shooting location of the target image, and to obtain target image inference data; the reference photovoltaic array spatial image model is constructed based on a reference photovoltaic array that is passable by obstacles affecting the path.

[0104] Optionally, the target image inference data includes first image inference data and second image inference data; the aforementioned judgment unit is specifically used to: determine whether the photovoltaic array at the shooting location has the connecting device based on the shooting coordinate information of the shooting location of the target image, and obtain the first image inference data; and determine whether the photovoltaic array at the shooting location has any obstacles other than the connecting device based on a preset reference photovoltaic array spatial image model and the target image, and obtain the second image inference data.

[0105] Optionally, the first waiting module 1120 includes: a waiting unit, used to pause the cleaning task if there is a traffic obstruction at the photovoltaic array at the shooting location, and wait for the drone to take a surround observation image taken by flying along a preset surround route; a modeling unit, used to acquire the surround observation image taken by the drone, construct a spatial image model of the surround observation image, and obtain an observation spatial image model; and a recognition unit, used to determine whether the traffic obstruction is passable through the observation spatial image model, and obtain a passability recognition result.

[0106] Optionally, the aforementioned identification unit is specifically used for: if the obstacle affecting passage is the connecting device, then based on the observation space image model, determining whether the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than a preset tilt angle difference threshold; if the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than the preset tilt angle difference threshold, then determining that the passageability identification result indicates that the obstacle affecting passage is not passable; if the tilt angle difference between the photovoltaic modules on both sides of the connecting device is less than or equal to the preset tilt angle difference threshold, then determining that the passageability identification result indicates that the obstacle affecting passage is passable.

[0107] The aforementioned photovoltaic cleaning robot's task execution device acquires image reasoning data derived from images captured by a drone during the cleaning process. It then determines whether to pause the cleaning task and waits for the accessibility recognition result before deciding whether to continue or stop the cleaning task. This allows for accurate determination of whether to continue the cleaning task based on whether the accessibility is obstructed, eliminating the need for human observation for accessibility recognition and reducing the labor costs of photovoltaic cleaning.

[0108] For the corresponding method embodiments described above, see [link to relevant documentation]. Figure 12The diagram illustrates a task instruction device for a photovoltaic cleaning robot, applicable to an unmanned aerial vehicle (UAV). The device includes: a second response module 1200, configured to respond to an inspection command and, based on the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot indicated by the inspection command, fly above the first photovoltaic cleaning robot to begin the cleaning task; wherein the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, with a maximum of one photovoltaic cleaning robot mounted on each row of modules; and a first imaging module 1210, configured to fly above the first photovoltaic cleaning robot along its direction of movement and capture images using a camera to obtain a target image, and to capture the target image while... The drone's location coordinates are captured; a second acquisition module 1220 is used to send the target image and the captured coordinates to a target processing device to obtain flight instructions generated based on the target image and the captured coordinates; the target processing device is used to instruct the photovoltaic cleaning robot or the cloud platform; a third response module 1230 is used for the drone to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next photovoltaic cleaning robot to start performing a cleaning task, in order to begin performing an inspection task for the second photovoltaic cleaning robot; a second stop module 1240 is used to stop performing all inspection tasks when all photovoltaic cleaning robots stop performing cleaning tasks.

[0109] The aforementioned task instruction device for photovoltaic cleaning robots allows drones to perform inspection tasks for one or more photovoltaic cleaning robots, resulting in low inspection costs. Furthermore, the macroscopic image data provided to the photovoltaic cleaning robots enables more accurate identification of their accessibility on the photovoltaic array.

[0110] For the corresponding method embodiments described above, see [link to relevant documentation]. Figure 13The diagram illustrates a task instruction device for a photovoltaic cleaning robot, applied to a cloud platform. The device includes: a first determining module 1300, used to determine the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in a photovoltaic array; wherein the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; each row of photovoltaic modules can mount at most one photovoltaic cleaning robot; a first sending module 1310, used to send a cleaning instruction to the first photovoltaic cleaning robot performing the cleaning task, and a patrol instruction to a drone, based on the starting coordinates and the cleaning sequence; the cleaning instruction instructs the first photovoltaic cleaning robot to begin performing the cleaning task, and the patrol instruction instructs the drone to begin performing a patrol task for the first photovoltaic cleaning robot; a first generating module 1310; and a first generating module 1310. Module 1320 is used to generate target image inference data based on the target image of the photovoltaic array captured by the UAV at its location and the shooting coordinate information of the shooting location of the target image; the target image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location; the second sending module 1330 is used to send an execution command to the first photovoltaic cleaning robot and a flight command to the UAV based on the target image inference data; the execution command is used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task; the flight command is used to instruct the UAV to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to start performing the cleaning task, so as to start performing an inspection task for the second photovoltaic cleaning robot.

[0111] Optionally, the first determining module 1300 is specifically used for: sending a modeling flight command to the drone to instruct the drone to fly along a preset modeling route and take modeling images; constructing a spatial image model of the photovoltaic array to be cleaned based on the modeling images to obtain a target photovoltaic array spatial image model; determining the starting coordinate information of the starting position of each photovoltaic cleaning robot in the photovoltaic array based on the target photovoltaic array spatial image model; and determining the cleaning sequence based on the starting coordinate information.

[0112] Optionally, the second sending module 1330 includes: a first sending submodule, configured to send a first execution command to the first photovoltaic cleaning robot and a first flight command to the drone if there are obstacles obstructing passage at the photovoltaic array at the shooting location, instructing the first photovoltaic cleaning robot to suspend the cleaning task and instructing the drone to start flying along a preset circumferential route at its current location and capture circumferential observation images; and a second sending submodule, configured to send a second execution command to the first photovoltaic cleaning robot and a second flight command to the drone if there are no obstacles obstructing passage at the photovoltaic array at the shooting location, instructing the first photovoltaic cleaning robot to continue the cleaning task and instructing the drone to continue flying along the direction of movement of the first photovoltaic cleaning robot.

[0113] Optionally, the second sending module 1330 further includes: an acquisition submodule for acquiring the surround observation image captured by the UAV; a construction submodule for constructing a spatial image model from the surround observation image to obtain an observation spatial image model; and a judgment submodule for determining whether the accessibility obstacle is passable based on the observation spatial image model to obtain a passability identification result.

[0114] Optionally, the second sending module 1330 further includes: a first indication submodule, configured to instruct the first photovoltaic cleaning robot to continue performing the cleaning task and the drone to continue flying along the movement direction of the first photovoltaic cleaning robot if the accessibility identification result indicates that the accessibility obstacle is passable; and a second indication submodule, configured to instruct the first photovoltaic cleaning robot to stop the cleaning task and the drone to fly above the next second photovoltaic cleaning robot to begin performing the cleaning task, if the accessibility identification result indicates that the accessibility obstacle is not passable.

[0115] The aforementioned task instruction device for the photovoltaic cleaning robot allows the cloud platform to uniformly assign tasks to both the photovoltaic cleaning robot and the drone, and to process complex image reasoning, generating image reasoning data to improve the execution and instruction efficiency of cleaning tasks.

[0116] This embodiment also provides an electronic device, including a processor and a memory. The memory stores machine-executable instructions that can be executed by the processor. The processor executes the machine-executable instructions to implement the above-described task execution method or task instruction method for the photovoltaic cleaning robot. This electronic device can be a server or a terminal device.

[0117] See Figure 14As shown, the electronic device includes a processor 1400 and a memory 1401. The memory 1401 stores machine-executable instructions that can be executed by the processor 1400. The processor 1400 executes the machine-executable instructions to implement the above-mentioned task execution method or task instruction method of the photovoltaic cleaning robot.

[0118] Furthermore, Figure 14 The electronic device shown also includes a bus 1402 and a communication interface 1403. The processor 1400, the communication interface 1403 and the memory 1401 are connected via the bus 1402.

[0119] The memory 1401 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 1403 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 1402 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 14 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0120] The processor 1400 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 1400 or by instructions in software form. The processor 1400 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 1401. Processor 1400 reads information from memory 1401 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0121] This embodiment also provides a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions cause the processor to implement the above-mentioned task execution method or task instruction method of the photovoltaic cleaning robot.

[0122] The computer program product of the photovoltaic cleaning robot task instruction and execution method and related equipment provided in this disclosure includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0123] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0124] Furthermore, in the description of the embodiments of this disclosure, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this disclosure based on the specific circumstances.

[0125] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0126] In the description of this disclosure, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this disclosure 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 disclosure. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0127] Finally, it should be noted that the above embodiments are merely specific implementations of this disclosure, used to illustrate the technical solutions of this disclosure, and not to limit it. The protection scope of this disclosure is not limited thereto. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this disclosure. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be covered within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the protection scope of the claims.

Claims

1. A task execution method for a photovoltaic cleaning robot, applied to a photovoltaic cleaning robot, characterized in that, The method includes: In response to a cleaning command, a cleaning task is initiated for the photovoltaic array, which comprises at least one row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device. During the cleaning task, target image inference data is acquired; the target image inference data is obtained based on the target image of the photovoltaic array taken by the drone at its location; the target image inference data includes the image inference result and the shooting coordinate information of the shooting location of the target image; the image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location; the obstacles affecting the passage include the connecting device; If there are obstacles affecting passage at the photovoltaic array at the shooting location, the cleaning task is paused and the passageability identification result of the obstacles is awaited. If the accessibility identification result indicates that the accessibility obstacle is passable, then the cleaning task continues; If the accessibility identification result indicates that the accessibility obstacle is impassable, then the cleaning task is stopped.

2. The method according to claim 1, characterized in that, The photovoltaic cleaning robot is connected to the drone for communication. The process of acquiring target image inference data during the execution of the cleaning task includes: During the cleaning task, target image inference data is generated based on the target image of the photovoltaic array taken by the drone and the shooting coordinate information of the target image's shooting position.

3. The method according to claim 2, characterized in that, During the cleaning task, target image inference data is generated based on the target image of the photovoltaic array captured by the drone and the shooting coordinates of the target image's shooting location, including: During the cleaning task, the drone captures a target image of the photovoltaic array from its location; the drone flies in the direction of movement of the photovoltaic cleaning robot and captures images periodically. Based on the preset reference photovoltaic array spatial image model, the target image, and the shooting coordinate information of the shooting position of the target image, it is determined whether there are any obstacles affecting the photovoltaic array at the shooting position, and target image inference data is obtained; the reference photovoltaic array spatial image model is constructed based on the reference photovoltaic array that is passable by the obstacles affecting the target.

4. The method according to claim 3, characterized in that, The target image inference data includes first image inference data and second image inference data; The step of determining whether there are any obstacles affecting the photovoltaic array at the shooting location based on a preset reference photovoltaic array spatial image model, the target image, and the shooting coordinates of the shooting location of the target image, and obtaining target image inference data, includes: Based on the shooting coordinates of the shooting location of the target image, it is determined whether the photovoltaic array at the shooting location has the connecting device, and the first image inference data is obtained; Based on the preset benchmark photovoltaic array spatial image model and the target image, it is determined whether there are any obstacles affecting traffic, other than the connecting device, at the photovoltaic array at the shooting location, and the second image inference data is obtained.

5. The method according to claim 1, characterized in that, If there are obstructions to traffic at the photovoltaic array at the shooting location, the cleaning task is paused, and the trafficability identification result of the obstructions is awaited, including: If there are obstacles obstructing the passage of the photovoltaic array at the shooting location, the cleaning task will be suspended, and the drone will be allowed to take a circumferential observation image while flying along a preset circumferential route. The drone captures surround observation images, and a spatial image model is constructed from the surround observation images to obtain an observation spatial image model. By using the observed spatial image model, it is determined whether the accessibility obstacle is passable, and a passability identification result is obtained.

6. The method according to claim 5, characterized in that, The step of determining whether the accessibility obstacle is passable through the observed spatial image model to obtain a passability identification result includes: If the obstruction to passage is the connecting device, then based on the observation space image model, it is determined whether the tilt angle difference between the photovoltaic modules on both sides of the connecting device is greater than a preset tilt angle difference threshold. If the tilt angle difference between the photovoltaic modules on both sides of the connection device is greater than the preset tilt angle difference threshold, then the accessibility identification result indicates that the accessibility obstruction is not passable; If the tilt angle difference between the photovoltaic modules on both sides of the connection device is less than or equal to a preset tilt angle difference threshold, then the accessibility identification result indicates that the accessibility obstacle is passable.

7. A task instruction method for a photovoltaic cleaning robot, applied to a drone, wherein the drone is equipped with a camera, characterized in that, The method includes: In response to the inspection command, according to the starting coordinate information and cleaning sequence of the starting position of at least one photovoltaic cleaning robot indicated by the inspection command, the robot flies to the top of the first photovoltaic cleaning robot that starts to perform the cleaning task, so as to start the inspection task for the first photovoltaic cleaning robot; wherein, the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and at most one photovoltaic cleaning robot is mounted on each row of photovoltaic modules. Above the first photovoltaic cleaning robot, the drone flies along the direction of movement of the first photovoltaic cleaning robot and captures images through the camera to obtain target images, as well as the shooting coordinate information of the drone's position when capturing the target images; The target image and the shooting coordinate information are sent to a target processing device to obtain flight commands generated based on the target image and shooting coordinate information; the target processing device is used to instruct the photovoltaic cleaning robot or cloud platform. According to the flight command, the drone starts flying along a preset circular route at its current location, continues to fly along the moving direction of the first photovoltaic cleaning robot, or flies above the second photovoltaic cleaning robot that starts to perform the cleaning task, so as to start performing the inspection task for the second photovoltaic cleaning robot. When all the photovoltaic cleaning robots stop performing cleaning tasks, all the inspection tasks will also stop.

8. A task instruction method for a photovoltaic cleaning robot, applied to a cloud platform, characterized in that, The method includes: The starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array are determined; wherein, the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; and each row of photovoltaic modules can carry at most one photovoltaic cleaning robot. Based on the starting coordinate information and the cleaning sequence, a cleaning instruction is sent to the first photovoltaic cleaning robot that first performs the cleaning task, and an inspection instruction is sent to the drone. The cleaning instruction is used to instruct the first photovoltaic cleaning robot to start performing the cleaning task, and the inspection instruction is used to instruct the drone to start performing an inspection task for the first photovoltaic cleaning robot. Based on the target image of the photovoltaic array taken by the UAV from its location, and the shooting coordinates of the target image's shooting location, target image inference data is generated; the target image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location. Based on the target image inference data, an execution command is sent to the first photovoltaic cleaning robot, and a flight command is sent to the drone. The execution command is used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task. The flight command is used to instruct the drone to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to begin performing a cleaning task, so as to begin performing an inspection task for the second photovoltaic cleaning robot.

9. The method according to claim 8, characterized in that, The determination of the starting coordinates and cleaning sequence of at least one photovoltaic cleaning robot in the photovoltaic array includes: Send modeling flight commands to the drone to instruct it to fly along a preset modeling route and capture modeling images; Based on the modeling image, a spatial image model of the photovoltaic array to be cleaned is constructed to obtain the spatial image model of the target photovoltaic array. Based on the spatial image model of the target photovoltaic array, determine the starting coordinate information of the starting position of each photovoltaic cleaning robot in the photovoltaic array; The cleaning sequence is determined based on the starting coordinate information.

10. The method according to claim 8, characterized in that, The step of sending execution commands to the first photovoltaic cleaning robot and flight commands to the drone based on the target image inference data includes: If there are obstacles obstructing the passage of the photovoltaic array at the shooting location, a first execution command is sent to the first photovoltaic cleaning robot and a first flight command is sent to the drone, instructing the first photovoltaic cleaning robot to suspend the cleaning task and instructing the drone to start flying along a preset circumferential route at its current location and take circumferential observation images. If there are no obstructions to the photovoltaic array at the shooting location, a second execution command is sent to the first photovoltaic cleaning robot, and a second flight command is sent to the drone, instructing the first photovoltaic cleaning robot to continue performing the cleaning task and instructing the drone to continue flying along the movement direction of the first photovoltaic cleaning robot.

11. The method according to claim 10, characterized in that, After sending a first execution command to the first photovoltaic cleaning robot and a first flight command to the drone if there is a traffic obstruction at the photovoltaic array at the shooting location, the method further includes: Acquire surround view images captured by the drone; A spatial image model is constructed from the surrounding observation image to obtain the observation spatial image model; By using the observed spatial image model, it is determined whether the accessibility obstacle is passable, and a passability identification result is obtained.

12. The method according to claim 11, characterized in that, After obtaining the accessibility identification result, the method further includes: If the accessibility identification result indicates that the accessibility obstacle is passable, then the first photovoltaic cleaning robot is instructed to continue performing the cleaning task, and the drone is instructed to continue flying along the movement direction of the first photovoltaic cleaning robot; If the accessibility identification result indicates that the accessibility obstacle is impassable, the first photovoltaic cleaning robot is instructed to stop the cleaning task, and the drone is instructed to fly over the second photovoltaic cleaning robot that will start the next cleaning task, so as to start the inspection task for the second photovoltaic cleaning robot.

13. A task execution method for a photovoltaic cleaning robot, applied to the task execution system of a photovoltaic cleaning robot, characterized in that, The system includes: a drone and at least one photovoltaic cleaning robot; the drone is equipped with a camera; the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and each row of photovoltaic modules carries at most one photovoltaic cleaning robot; the passage gaps between each row of photovoltaic modules are connected by a connecting device; The first photovoltaic cleaning robot responds to the cleaning command and begins to perform the cleaning task for the photovoltaic array. The first photovoltaic cleaning robot is the first photovoltaic cleaning robot to perform the cleaning task. The drone responds to the inspection command and begins to perform an inspection task on the first photovoltaic cleaning robot; The drone flies above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and takes pictures through the camera to obtain the target image and the shooting coordinate information of the drone's position when taking the target image; During the execution of the cleaning task, the first photovoltaic cleaning robot acquires target image inference data; the target image inference data is obtained based on the target image; the target image inference data includes image inference results and the shooting coordinate information; the image inference results are used to indicate whether there are any obstacles affecting the passage of the photovoltaic array at the shooting location of the target image; the obstacles affecting the passage include the connecting device; If there are obstacles affecting passage at the photovoltaic array at the shooting location, the first photovoltaic cleaning robot will pause the cleaning task and wait for the passage recognition result of the obstacles. The drone begins to fly along a preset circling route from its current location and takes circling observation images; The first photovoltaic cleaning robot acquires the accessibility recognition result generated based on the surrounding observation image, and continues or stops performing the cleaning task based on the accessibility recognition result; Based on the accessibility recognition result, the drone continues to perform the inspection task or begins to perform the inspection task for the second photovoltaic cleaning robot. When all photovoltaic cleaning robots stop performing cleaning tasks, the drone stops performing all inspection tasks.

14. A task execution device for a photovoltaic cleaning robot, applied to a photovoltaic cleaning robot, characterized in that, The device includes: The first response module is used to respond to a cleaning command and begin to execute a cleaning task for the photovoltaic array; the photovoltaic array includes at least one row of photovoltaic modules; the passage gaps between each row of photovoltaic modules are connected by a connecting device; The first acquisition module is used to acquire target image inference data during the execution of the cleaning task; the target image inference data is obtained by inferring from the target image of the photovoltaic array taken by the drone at its location; the target image inference data includes image inference results and shooting coordinate information of the shooting location of the target image; the image inference results are used to indicate whether there are any obstacles affecting the passage of the photovoltaic array at the shooting location; the obstacles affecting the passage include the connecting device; The first waiting module is used to pause the cleaning task and wait for the accessibility identification result of the photovoltaic array at the shooting location if there is an obstacle affecting passage. The first execution module is configured to continue executing the cleaning task if the accessibility identification result indicates that the accessibility obstacle is passable. The first stop module is used to stop the cleaning task if the accessibility identification result indicates that the accessibility-affecting object is impassable.

15. A task instruction device for a photovoltaic cleaning robot, applied to a drone, the drone being equipped with a camera, characterized in that, The device includes: The second response module is used to respond to the inspection command and, according to the starting coordinate information and cleaning sequence of the starting position of at least one photovoltaic cleaning robot indicated by the inspection command, fly to above the first photovoltaic cleaning robot that starts to perform the cleaning task, so as to start to perform the inspection task for the first photovoltaic cleaning robot; wherein, the photovoltaic array to be cleaned includes at least one row of photovoltaic modules, and at most one photovoltaic cleaning robot is mounted on each row of photovoltaic modules; The first shooting module is used to fly above the first photovoltaic cleaning robot along the direction of movement of the first photovoltaic cleaning robot, and to take pictures through the camera to obtain a target image, as well as the shooting coordinate information of the drone's position when taking the target image; The second acquisition module is used to send the target image and the shooting coordinate information to the target processing device to obtain flight commands generated based on the target image and the shooting coordinate information; the target processing device is used to instruct the photovoltaic cleaning robot or the cloud platform. The third response module is used for the UAV to start flying along a preset circular route at its current location, continue flying along the moving direction of the first photovoltaic cleaning robot, or fly above the second photovoltaic cleaning robot that starts to perform the cleaning task, in order to start performing the inspection task for the second photovoltaic cleaning robot, according to the flight command. The second stop module is used to stop all the inspection tasks when all the photovoltaic cleaning robots stop performing cleaning tasks.

16. A task instruction device for a photovoltaic cleaning robot, applied to a cloud platform, characterized in that, The device includes: The first determining module is used to determine the starting coordinate information and cleaning sequence of the starting position of at least one photovoltaic cleaning robot in the photovoltaic array; wherein, the cleaning sequence is used to indicate the starting order in which the photovoltaic cleaning robots perform cleaning tasks; the photovoltaic array includes at least one row of photovoltaic modules; and each row of photovoltaic modules can be equipped with at most one photovoltaic cleaning robot; The first sending module is used to send a cleaning instruction to the first photovoltaic cleaning robot that first performs the cleaning task, and to send an inspection instruction to the drone, according to the starting coordinate information and the cleaning sequence. The cleaning instruction is used to instruct the first photovoltaic cleaning robot to start performing the cleaning task, and the inspection instruction is used to instruct the drone to start performing an inspection task for the first photovoltaic cleaning robot. The first generation module is used to generate target image inference data based on the target image of the photovoltaic array taken by the UAV at its location and the shooting coordinate information of the shooting location of the target image; the target image inference result is used to indicate whether there are any obstacles affecting the photovoltaic array at the shooting location. The second sending module is used to send execution instructions to the first photovoltaic cleaning robot and flight instructions to the drone based on the target image inference data. The execution instructions are used to instruct the first photovoltaic cleaning robot to pause, continue, or stop performing the cleaning task. The flight instructions are used to instruct the drone to start flying along a preset circular route at its current location, continue flying along the movement direction of the first photovoltaic cleaning robot, or fly above the next second photovoltaic cleaning robot to begin performing a cleaning task, so as to start performing an inspection task for the second photovoltaic cleaning robot.

17. An electronic device, characterized in that, It includes a processor and a memory, the memory storing machine-executable instructions that can be executed by the processor, the processor executing the machine-executable instructions to implement the task execution method of the photovoltaic cleaning robot or the task instruction method of the photovoltaic cleaning robot according to any one of claims 1-13.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the task execution method of the photovoltaic cleaning robot or the task instruction method of the photovoltaic cleaning robot according to any one of claims 1-13.

Images