Path planning method of photovoltaic robot, photovoltaic robot, equipment and medium

Abstract

The invention relates to a path planning method of a photovoltaic robot, the photovoltaic robot, equipment and a medium, and the method comprises the steps: setting a distance between adjacent paths according to a machine cleaning performance index, generating an internal cleaning path based on a photovoltaic plate map and the distance between the adjacent paths, and discretizing the internal cleaning path into a plurality of path points with expected poses; an acceptance path point is arranged on each sub-cleaning path of the internal cleaning path, and parameters of the acceptance path points comprise an expected cleaning area; executing an internal cleaning path, monitoring a pose error between an actual pose and an expected pose in real time, and determining a cleaning performance error according to an actual cleaning area and an expected cleaning area; when the pose error is greater than the pose tolerance threshold or the cleaning performance error is less than the cleaning tolerance threshold, a target internal cleaning path is generated and executed based on the pose error and / or the cleaning performance error. The execution speed and the cleaning efficiency of the photovoltaic robot on the cleaning path are improved, and meanwhile it is ensured that the cleaning quality reaches the expected standard.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic panel cleaning, and in particular to a path planning method of a photovoltaic robot, a photovoltaic robot, equipment and a medium. Background Art

[0002] Solar photovoltaic panels, also known as solar panels, are devices that convert sunlight directly into electrical energy. In order to optimize the sunlight receiving efficiency of solar photovoltaic panels, they are usually installed on structures such as roofs at an angle. However, long-term outdoor exposure will lead to the accumulation of dust and dirt, which will significantly reduce the light energy conversion efficiency of photovoltaic panels. Therefore, regular cleaning of photovoltaic panels is essential to maintain their high performance. Photovoltaic robots can effectively improve the efficiency of removing dust, dirt and snow from the surface of photovoltaic panels through automated cleaning operations, avoiding the high cost and risk of damage to photovoltaic panels and high-altitude operations caused by manual operations.

[0003] In the diverse photovoltaic power station environment, photovoltaic robots face multiple challenges, such as adapting to photovoltaic panels of different sizes, tilted photovoltaic panel cleaning surfaces, etc. In order to achieve accurate path planning, the robot relies on advanced perception modules, which collect environmental data to provide reliability of the cleaning path.

[0004] For example, the Chinese invention patent with application publication number CN117032266A discloses a photovoltaic cleaning robot driven by a brushless motor and its path planning method, which rotates the photovoltaic panel map coordinate system constructed based on the satellite positioning signal, converts the bottom edge of the photovoltaic panel in the photovoltaic panel map to the X-axis of the positioning coordinate system, establishes a mapping relationship between each coordinate position in the photovoltaic panel map and the positioning coordinate system, and thus constructs the traversal path of the photovoltaic cleaning robot in the positioning coordinate system according to the rotated photovoltaic panel map. Each coordinate point in the path thus obtained can correspond one-to-one to the physical coordinate position in the photovoltaic map, and the traversal path of the photovoltaic cleaning robot can be quickly and completely planned directly through a general traversal algorithm. (Safety index + adjustment posture). The Chinese invention patent with application publication number CN11644809A discloses a photovoltaic robot positioning and navigation method and device based on SLAM technology. The method includes: obtaining an environmental point cloud map of the photovoltaic robot through a rotatable laser radar installed on the photovoltaic robot; obtaining a scene image of the photovoltaic robot through a visual camera installed on the photovoltaic robot; fusing the environmental point cloud map and the scene image at the same time, using SLAM to build a map, and obtaining an omnidirectional three-dimensional photovoltaic site map; based on the omnidirectional three-dimensional photovoltaic site map, combined with an obstacle avoidance sensor installed on the photovoltaic robot, positioning, identifying and navigating the photovoltaic robot, and controlling the photovoltaic robot to reach a destination. The Chinese invention patent with application publication number CN114877892 A discloses a fusion positioning method for a photovoltaic robot, which belongs to the field of robot vision technology, including: obtaining relevant data sources in the movement of the photovoltaic robot through multiple data source acquisition devices installed on the photovoltaic robot; processing part of the data in the obtained relevant data sources to obtain new data; establishing an observation model based on an extended Kalman filter; inputting part of the relevant data sources in the movement of the photovoltaic robot and the new data obtained by processing part of the data in the obtained relevant data sources into the observation model to obtain multiple observation quantities; adjusting relevant parameters of the extended Kalman filter; obtaining the final fusion positioning data of the photovoltaic robot. The present invention inputs the processed GPS data and INS data into the observation model for fusion, which can avoid the situation where INS data is saturated and missing for a short time due to robot vibration, and GPS positioning failure caused by occlusion and the like.

[0005] However, in complex and diverse environments, photovoltaic robots are still not flexible enough when cleaning photovoltaic arrays, and are prone to problems such as incomplete coverage and excessive energy consumption. Summary of the invention

[0006] The main purpose of the present invention is to provide a path planning method for a photovoltaic robot, a photovoltaic robot, a device and a medium. In order to solve the above-mentioned technical problems, the present invention specifically adopts the following technical solutions: A first aspect of the present invention is to provide a path planning method for a photovoltaic robot, wherein the photovoltaic robot comprises at least two perception modules, wherein the perception modules comprise a visual perception module, and the method comprises: S201: obtaining a machine cleaning performance index, and setting an adjacent path spacing between adjacent sub-cleaning paths according to the machine cleaning performance index; S202: obtaining a photovoltaic panel map, generating an internal cleaning path based on the photovoltaic panel map and the adjacent path spacing, and discretizing the internal cleaning path into a plurality of path points, wherein the parameters of the path points include an expected posture; S203 setting an acceptance path point on each sub-cleaning path of the internal cleaning path, wherein parameters of the acceptance path point include an expected cleaning area; S204 controls the photovoltaic robot to execute the internal cleaning path; monitors in real time the posture error between the actual posture acquired by the perception module and the expected posture; obtains the actual cleaning area based on the visual perception module at each acceptance path point, and determines the cleaning performance error according to the actual cleaning area and the expected cleaning area; S205: When the posture error is greater than the posture tolerance threshold or the cleaning performance error is greater than the cleaning tolerance threshold, a target internal cleaning path is generated based on the posture error and / or the cleaning performance error, and the photovoltaic robot is controlled to execute the target internal cleaning path.

[0007] In some embodiments, the method further includes: when the posture error is less than a posture tolerance threshold and the cleaning performance error is less than a cleanliness tolerance threshold, controlling the photovoltaic robot to continue executing the internal cleaning path.

[0008] In some embodiments, the target internal cleaning path includes a posture correction path and an optimized cleaning path, and S205 includes: generating a posture correction path of the photovoltaic robot based on the posture error; and / or updating a machine cleaning performance index based on the cleaning performance error; updating an adjacent path spacing between adjacent sub-cleaning paths according to the updated machine cleaning performance index; generating the optimized cleaning path based on the updated adjacent path spacing, and controlling the photovoltaic robot to execute the optimized cleaning path.

[0009] In some embodiments, the S205 also includes: controlling the photovoltaic robot to move according to the posture correction path; obtaining a corrected posture based on the perception module, and if the posture error between the corrected posture and the expected posture is less than a preset correction error threshold, the posture correction is successful; controlling the photovoltaic robot to execute the optimized cleaning path.

[0010] In some embodiments, the S204 also includes: when the photovoltaic robot reaches the acceptance path point, obtaining the image acceptance data collected by the visual perception module; based on a preset recognition model, identifying the final cleaning area image in the image acceptance data; and calculating the actual cleaning area corresponding to the final cleaning area image based on a preset mapping algorithm.

[0011] In some embodiments, the cleaning performance index includes a cleaning force index and a cleaning coverage index, and S201 also includes: determining the minimum number of repeated cleaning times based on the cleaning force index and the expected cleaning standard; determining the adjacent path spacing between adjacent sub-cleaning paths based on the cleaning coverage index and the minimum number of repeated cleaning times.

[0012] In some embodiments, the method also includes: obtaining the fluctuation frequency of the actual posture; when the fluctuation frequency is not within a preset frequency range, identifying whether there is a high-risk area on the cleaning path based on the environmental data collected by the perception module, and if there is no high-risk area on the cleaning path, controlling the photovoltaic robot to continue to execute the internal cleaning path; if there is a high-risk area on the cleaning path, controlling the photovoltaic robot to bypass the high-risk area and continue to execute the internal cleaning path.

[0013] A second aspect of the present invention is to provide a photovoltaic robot, the photovoltaic robot comprising: A moving device for moving between photovoltaic panels; A detection device, installed on the mobile device, the detection device includes at least two perception modules, and the perception modules are used to collect motion data or environmental data of the photovoltaic robot; A cleaning device, installed on the mobile device, for cleaning the photovoltaic panel; A control system is connected to the mobile device, the detection device and the cleaning device, and is used to implement the steps of the path planning method for the photovoltaic robot provided in any embodiment of the present invention.

[0014] A third aspect of the present invention is to provide a computer device, the computer device comprising a memory and a processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the steps of the path planning method for a photovoltaic robot provided in any embodiment of the present invention when executing the computer program.

[0015] The fourth aspect of the present invention is to provide a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor performs the steps of the path planning method for a photovoltaic robot provided in any embodiment of the present invention.

[0016] Beneficial technical effects: The present invention provides a path planning method for a highly adaptable photovoltaic robot that improves cleaning efficiency while ensuring cleaning quality. Various photovoltaic robots with different cleaning ranges and cleaning capabilities are adapted to flexibly determine the spacing between adjacent paths, and customized internal cleaning paths that meet the expected cleaning effects are generated to ensure that the planning of the cleaning path can be scalably and accurately adapted to the needs of different cleaning tasks. Furthermore, during the photovoltaic panel cleaning process, the execution of the cleaning task is monitored in real time from two dimensions: path deviation and photovoltaic panel cleaning effect, and different processing modes are adopted for different degrees of path execution errors.

[0017] The first is the slow response mode. When cleaning the internal area of ​​the photovoltaic panel array, the movement of the photovoltaic robot in this area is relatively safe, and in order to ensure the cleaning quality, some areas will be cleaned repeatedly. As a result, the photovoltaic robot enjoys a certain degree of freedom in the execution of the path and can tolerate path execution errors that do not affect the cleaning quality and travel safety, avoid frequent adjustments to the cleaning path, reduce unnecessary repetitions of the cleaning path, improve the photovoltaic robot's execution speed and cleaning efficiency of the cleaning path, and save energy and time.

[0018] The second is the rapid response mode. When the cleaning quality and travel safety cannot be effectively guaranteed, the position posture and path spacing are adjusted in time according to the real-time data to improve the effectiveness and flexibility of the cleaning task. For example, the posture correction path is generated through the real-time monitoring of posture errors to improve the reliability and safety of the operation; the spacing between adjacent paths is adaptively adjusted through the cleaning results of fixed-point monitoring to generate an optimized cleaning path to ensure that the cleaning work meets the expected standards, improve the overall cleaning quality, and thus improve the sustainable execution capability of the cleaning task.

[0019] Furthermore, the fluctuation frequency of the actual posture is evaluated to identify sudden position changes caused by high-risk areas (such as damaged areas and slipping areas), and the cleaning path is flexibly adjusted. Additional monitoring is performed on abnormal fluctuations in data within the tolerance range of path execution errors to adapt to different environmental conditions and photovoltaic panel conditions, thereby improving the adaptability, safety and reliability of the photovoltaic robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, each element or part is not necessarily drawn according to the actual scale. Obviously, the drawings described below are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without paying creative labor.

[0021] Figure 1 Schematic diagram of two array arrangements of photovoltaic panels provided by embodiments of the present invention; Figure 2 is a schematic diagram of a photovoltaic panel map provided by an embodiment of the present invention; Figure 3 is a schematic flow chart of a photovoltaic array cleaning method provided by an embodiment of the present invention; Figure 4 is a schematic diagram of a cleaning path of a photovoltaic robot provided by an embodiment of the present invention; Figure 5 is a partial schematic diagram of a cleaning path of a photovoltaic robot provided by an embodiment of the present invention; Figure 6 is a schematic flow chart of a path planning method for a photovoltaic robot provided in an embodiment of the present invention; Figure 7 is a schematic diagram of an acceptance path point provided by an embodiment of the present invention; Figure 8 It is a schematic block diagram of the structure of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0023] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0024] Herein, suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention, and have no specific meanings by themselves. Therefore, "module", "component" or "unit" can be used mixedly.

[0025] In this document, the terms "upper", "lower", "inner", "outer", "front", "back", "one end", "the other end" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0026] In this document, unless otherwise clearly specified and limited, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0027] Herein "and / or" includes any and all combinations of one or more of the associated listed items.

[0028] Herein, "plurality" means two or more than two, ie, it includes two, three, four, five, etc.

[0029] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0030] In order to adapt to the complex environment of photovoltaic power stations, photovoltaic robots need to call a large number of perception modules during the cleaning process to collect and process a large amount of data in real time, and use these data to ensure that they are on the planned path to ensure that the photovoltaic robots clean the photovoltaic panels according to the planned path. This process will consume a lot of electricity.

[0031] Furthermore, when the photovoltaic robot deviates from the path, it needs to be corrected immediately to the planned path to ensure the safety and cleaning coverage of the photovoltaic robot's high-altitude operation. Frequent posture orientation and position correction further aggravates the energy consumption problem and reduces the efficiency of path execution. Among them, the posture orientation and position are collectively referred to as posture below.

[0032] Therefore, under the premise of ensuring safety, the present invention optimizes the cleaning strategy of the photovoltaic panel array and the path planning algorithm of the photovoltaic robot, refines the energy consumption management of the photovoltaic robot, improves the flexibility, adaptability and energy efficiency of the photovoltaic robot, and ensures that it can stably and efficiently complete the cleaning task in various complex environments. This will not only help improve the operation and maintenance level of photovoltaic power stations, but will also promote the further development and application of photovoltaic technology.

[0033] A photovoltaic array is a DC power generation unit with a fixed support structure, which is composed of multiple photovoltaic modules or photovoltaic panels mechanically and electrically assembled together. The photovoltaic panel map in the embodiment of the present invention is a photovoltaic panel layout map or a photovoltaic panel array map, which is a map used to record the arrangement and layout of photovoltaic panels.

[0034] In some embodiments, photovoltaic panel layout information input by a user is obtained, and a photovoltaic panel map is generated according to the photovoltaic panel layout information. The photovoltaic panel layout information is a key parameter in the installation of a photovoltaic panel array, which may be photovoltaic panel size information, panel tilt information, inter-panel gap information, array arrangement information, etc.

[0035] Specifically, the photovoltaic panel size information refers to the physical size of the photovoltaic panel, which may include length, width and thickness; the panel tilt information refers to the angle between the photovoltaic panel and the horizontal plane, that is, the installation tilt angle; the inter-panel gap information refers to the size of the gap between the photovoltaic panels. The array arrangement information refers to the arrangement of the photovoltaic panels in the photovoltaic panel array.

[0036] For example, the array arrangement information includes the distribution direction of the photovoltaic panels, see Figure 1 , Figure 1 Schematic diagram of two array arrangements of photovoltaic panels provided by embodiments of the present invention, such as Figure 1 As shown, photovoltaic panels can be placed vertically or horizontally.

[0037] For example, if the user inputs the photovoltaic panel size information "the height of the photovoltaic panel is 1600mm, the width of the photovoltaic panel is 800mm", the inter-panel gap information "the gap between two photovoltaic panels is 40mm", and the array arrangement information "the vertical arrangement of 4 rows and 16 columns of photovoltaic panels", please refer to Figure 2 , Figure 2 is a schematic diagram of a photovoltaic panel map provided by an embodiment of the present invention. According to the photovoltaic panel layout information in this example, the following can be mapped and generated: Figure 2Photovoltaic panel map shown.

[0038] In some embodiments, the photovoltaic robot includes at least two perception modules, which are used to enable the photovoltaic robot to perceive the surrounding environment, collect motion data and environmental data of the photovoltaic robot, and then understand the actual position of the fuselage and the cleaning status of the photovoltaic panels. The perception modules can be various visual perception modules, speed perception modules, displacement perception modules, etc. Specifically, they can be wheel odometers, inertial measurement units, visual sensors, ultrasonic sensors, environmental sensors, satellite positioning systems, laser radars, etc.

[0039] The present invention proposes a highly adaptable photovoltaic array cleaning method that takes into account both cleaning effect and endurance. Based on the algorithm-generated internal cleaning path, the boundary compensation cleaning path is extended to explore, and the cleaning strategy for the internal area and boundary area of ​​the photovoltaic panel array is adaptively customized. All perception modules are used at the first boundary cleaning station in the trust spacing area for precise boundary exploration, and some perception modules are used for simple edge exploration at other boundary cleaning stations. By distinguishing between the main edge exploration cleaning station and the slave edge exploration cleaning station, a differentiated perception module scheduling strategy is implemented. While ensuring high coverage and high cleanliness of photovoltaic panel cleaning, the repetitive work in the edge exploration process is reduced, energy is saved, and consumption is reduced, the resource utilization of photovoltaic robots is optimized, and the cleaning efficiency and sustainable operation capability of photovoltaic robots are effectively improved, ensuring the long-term stable operation of photovoltaic power stations and maximizing energy output.

[0040] See also Figure 3 , Figure 3 is a schematic flow chart of a photovoltaic array cleaning method provided by an embodiment of the present invention, such as Figure 3 As shown, the method is applied to a photovoltaic robot, including steps S101 to S104.

[0041] S101 obtains a photovoltaic panel map, and generates an internal cleaning path based on the photovoltaic panel map, wherein the internal cleaning path passes through a plurality of cleaning stations, and the cleaning stations include boundary cleaning stations with a preset distance value from any boundary to be cleaned of the photovoltaic panel map.

[0042] The internal cleaning path includes multiple sub-cleaning paths, each of which covers the internal area relatively far from the boundary in the photovoltaic panel map. In order to make the internal cleaning path fully cover the internal area, the internal cleaning path passes through several cleaning stations, which are the speed change points or turning points of the photovoltaic robot in the internal cleaning path. Correspondingly, the boundary cleaning station is a cleaning station that changes speed or turns near any boundary to be cleaned in the photovoltaic panel map.

[0043] Specifically, the photovoltaic panel map is divided into an internal area and a boundary area by a preset distance value, and then the cleaning strategy for cleaning the internal area and the boundary area of ​​the photovoltaic panel array is adaptively customized. The specific value of the preset distance value can be flexibly determined according to actual needs. Exemplarily, the preset distance value can be 50cm, and a virtual internal area boundary is formed at a distance of 50cm from the boundary. The area within the virtual internal area boundary is the internal area, and the area outside the virtual internal area boundary is the boundary area. Then, based on the path planning algorithm, an internal cleaning path is generated for the internal area of ​​the photovoltaic panel map. Correspondingly, the cleaning station on the boundary of the virtual internal area is the boundary cleaning station. The path planning algorithm can refer to the relevant technology and is not limited here.

[0044] It should be understood that when the photovoltaic robot executes the internal cleaning path, it can only clean the internal area. In this area, the photovoltaic robot is far away from the boundary, there is no risk of falling from a high altitude, and the movement is relatively safe. At the same time, the photovoltaic panels in the internal area are installed more neatly and regularly than those in the boundary, there are fewer obstacles, and the corresponding map is more accurate. Therefore, the internal area excludes most of the complex interference factors in the environment, and the path planning algorithm used in this area has higher reliability and practicality.

[0045] S102 controls the photovoltaic robot to execute the internal cleaning path, and when reaching the boundary cleaning station, detects whether there is a main edge detection cleaning station of the same boundary to be cleaned in the trust distance area.

[0046] The photovoltaic robot performs the cleaning task according to the internal cleaning path to complete the cleaning of the internal area. When the photovoltaic robot reaches the boundary cleaning station, it cleans the boundary area. First, the boundary to be cleaned within the preset distance value of the boundary cleaning station is determined, and then the boundary to be cleaned is detected to determine whether there is a main edge detection cleaning station in the trust interval area of ​​the boundary cleaning station to determine the perception module scheduling strategy of the boundary cleaning station.

[0047] The trust distance area is the area that is extended from the current boundary cleaning station along the same boundary to the cleaned area by the preset trust distance. The preset trust distance can be flexibly set according to the actual situation. For example, the trust distance area can be set to 1m, that is, the area that is extended from the boundary cleaning station along the same boundary to the cleaned area by 1m. It should be noted that the trust distance area is the cleaned area, and the main edge cleaning station is found in the path executed by the photovoltaic robot, which improves the reliability of the reference cleaning path.

[0048] Exemplarily, the trust interval area is the length area or width area of ​​a single photovoltaic panel laid out on the photovoltaic panel map. Since the shape of the photovoltaic panel is a regular rectangle, according to the arrangement of the photovoltaic panel array, the long side or wide side of the rectangle is determined as the trust interval area, such as Figure 2As shown, when the photovoltaic panels are arranged vertically, the trust spacing area is the length area of ​​a single photovoltaic panel.

[0049] It should be understood that the trust spacing area is set based on the structural characteristics and laying rules of the photovoltaic panels in the photovoltaic panel map. The multiple boundary cleaning stations in this spacing area have a high degree of similarity with the boundary cleaning path of the boundary to be cleaned. Therefore, multiple slave edge detection cleaning stations close to the same boundary in the trust spacing area can trust the reference cleaning path of the main edge detection cleaning station for simple edge detection. That is, only one deep edge detection cleaning can be performed, and other boundary cleaning stations can quickly, accurately and low-energy complete the cleaning of the boundary area by referring to the deep edge detection data, which improves the safety of simple edge detection in the boundary cleaning of the photovoltaic cleaning machine.

[0050] S103: If there is no main edge detection cleaning station for the same boundary to be cleaned, the boundary cleaning station is set as the main edge detection cleaning station, all the perception modules are enabled to perform deep edge detection on the boundary to be cleaned, and a reference cleaning path for the main edge detection cleaning station is generated.

[0051] Specifically, if there is no main edge detection cleaning station in the trust spacing area, the current boundary cleaning station is set as the main edge detection cleaning station, and all perception modules are enabled to perform deep edge detection on the boundary to be cleaned, to achieve high-precision boundary recognition and posture speed control, to ensure full coverage and cleaning of the boundary during the edge detection process, while ensuring the safety of the photovoltaic robot, and to generate a reference cleaning path based on the deep edge detection data. It should be understood that there are multiple reference cleaning paths for the boundary to be cleaned, and each reference cleaning path is only effective when the boundary to be cleaned is within a certain area, that is, when arriving at the boundary cleaning station, when the main edge detection cleaning station of the same boundary to be cleaned is detected in the trust spacing area, the reference cleaning path of the main edge detection cleaning station is effective.

[0052] It should be noted that all the perception modules enabled in this embodiment are all perception modules used for boundary recognition and posture and speed control.

[0053] In some embodiments, S103 includes: controlling the photovoltaic robot to move toward the boundary to be cleaned at a preset edge detection value based on a first speed, wherein the preset edge detection value is greater than the preset distance value; enabling all of the perception modules to collect multiple environmental data in real time, and performing boundary identification based on the multiple environmental data, and when the boundary is identified, controlling the photovoltaic robot to stop moving; obtaining the moving path of the photovoltaic robot as a reference cleaning path for the main edge detection cleaning station.

[0054] Among them, the first speed is determined according to the remaining distance between the photovoltaic robot and the boundary when it recognizes the boundary, and the buffer distance required for the photovoltaic robot to brake. When the photovoltaic robot moves at the first speed and detects the boundary, it can stop before reaching the edge safety distance. The specific speed value can be flexibly set according to actual needs and is not limited here. For example, if the remaining distance is 18cm, the buffer distance is 10cm, and the edge safety distance is 8cm, then when the photovoltaic robot moves at the first speed, it can stop when it is 8cm away from the edge of the boundary to ensure the safety of the photovoltaic robot.

[0055] The preset edge detection value is greater than the preset distance value, so that when the photovoltaic panel map does not match the actual photovoltaic panel array, the map anomaly can be identified, and the photovoltaic panels can be fully cleaned under the condition of ensuring safety, thereby improving the coverage and cleanliness of the photovoltaic panel cleaning. For example, the preset edge detection value is set to 100cm and the preset distance value is 50cm. Since all perception modules are enabled during deep edge detection, the boundary can be accurately detected, allowing the photovoltaic robot to freely explore within a limited range of 50cm.

[0056] In some embodiments, the boundary detection of the depth edge detection can be performed by the visual perception module through image recognition boundary and the infrared sensor perception module through cliff detection boundary recognition.

[0057] S104: if there is a main edge detection cleaning station for the same boundary to be cleaned, the boundary cleaning station is set as a secondary edge detection cleaning station, the photovoltaic robot is controlled to obtain and execute the reference cleaning path of the main edge detection cleaning station, and part of the perception module is enabled to perform simple edge detection on the boundary to be cleaned.

[0058] Specifically, if there is a main edge detection cleaning station in the trust distance area, the current boundary cleaning station is set as a slave edge detection cleaning station, and the photovoltaic robot is controlled to execute the reference cleaning path of the main edge detection cleaning station to reduce repetitive work in the edge detection process, and some perception modules are enabled to perform simple edge detection on the boundary to be cleaned to avoid safety risks caused by local damage to the photovoltaic panels, and achieve high-precision, low-risk, short-time, and low-power efficient boundary cleaning.

[0059] It should be understood that there is at most one main edge detection cleaning station in the trust distance area. When the trust distance area is 1m, if the first boundary cleaning station detects that there is no main edge detection cleaning station within 1m of the same boundary to be cleaned, the first boundary cleaning station will be set as the first main edge detection cleaning station, and correspondingly, the second boundary cleaning station of the same boundary within 1m will detect the first main edge detection cleaning station and be set as a secondary edge detection cleaning station. Furthermore, if the photovoltaic robot reaches the third boundary cleaning station 1m away, the first main edge detection cleaning station is not within the trust distance area of ​​the third boundary cleaning station, therefore, the third boundary cleaning station is set as the second main edge detection cleaning station.

[0060] In some embodiments, after the simple edge detection is completed at the edge detection cleaning station, an actual cleaning path of the edge detection cleaning station is generated.

[0061] It should be noted that the partial perception modules enabled in this embodiment are partial perception modules used for boundary recognition and posture speed control. One or more can be selected according to actual needs. The purpose is to achieve energy saving and consumption saving by only using specific perception modules during simple edge detection, optimize the resource utilization of the photovoltaic robot, and effectively improve the sustainable operation capability of the photovoltaic robot.

[0062] In some embodiments, the S104 also includes: controlling the photovoltaic robot to execute the reference cleaning path based on the second speed, enabling any perception module to collect environmental data in real time, and performing boundary identification based on the environmental data; when the boundary is identified and / or the reference cleaning path is completed, controlling the photovoltaic robot to stop moving.

[0063] The second speed is determined based on the remaining distance between the photovoltaic robot and the boundary when the photovoltaic robot recognizes the boundary and the buffer distance required for the photovoltaic robot to brake. The second speed can be greater than or equal to the first speed. When the photovoltaic robot moves at the second speed and detects the boundary, it can at least stop before reaching the edge. The specific speed value can be flexibly set according to actual needs and is not limited here.

[0064] For example, if the remaining distance is 18 cm and the buffer distance is 10 cm, the photovoltaic robot can at least stop when it is close to the edge of the boundary when traveling at the second speed, thereby ensuring the safety of the photovoltaic robot.

[0065] In some embodiments, the photovoltaic robot moves at a third speed when executing the internal cleaning path, and the first speed and the second speed are both less than the third speed, for example, the third speed is 0.4 m / s, the first speed is 0.1 m / s, and the second speed is 0.2 m / s. By adapting different cleaning paths to different speeds, the cleaning efficiency of the photovoltaic robot is effectively improved while ensuring the safety of the photovoltaic robot.

[0066] In some embodiments, the boundary detection of the simple edge detection can be performed by the visual perception module through image recognition of the boundary and the infrared sensor perception module through cliff detection to recognize the boundary.

[0067] Therefore, by distinguishing between the main edge cleaning station and the slave edge cleaning station, a differentiated perception module scheduling strategy is implemented, the cleaning path is optimized, and repetitive work is reduced. The photovoltaic robot can dynamically adjust the cleaning path according to the actual environment and cleaning needs to ensure the efficiency and coverage of the cleaning work.

[0068] See also Figures 4 to 5 , Figure 4 is a schematic diagram of a cleaning path of a photovoltaic robot provided by an embodiment of the present invention, Figure 5 FIG. 1 is a partial schematic diagram of a cleaning path of a photovoltaic robot provided by an embodiment of the present invention. Figure 4 , Figure 5 As shown, the dotted line is the cleaning path, including an internal cleaning path of the internal area and a boundary cleaning path of the boundary area, wherein the internal cleaning path includes multiple sub-cleaning paths and multiple transition paths connecting the multiple sub-cleaning paths, and the boundary cleaning path includes a reference cleaning path of the main edge detection cleaning station.

[0069] In addition, the boundary cleaning path may also include an actual cleaning path from the edge detection cleaning station, a cleaning path of an independent edge detection station, and a verification cleaning path of a verification edge detection station, which are not shown in the figure.

[0070] The embodiment of the present invention also provides an intelligent switching mechanism for the edge detection strategy to cope with the diverse boundary scenarios caused by complex photovoltaic panel layouts and photovoltaic panel damage. While simplifying the use of the perception module, it avoids the security issues caused by data uniformity or accuracy limitations, thereby improving the safety of simple edge detection in boundary cleaning of photovoltaic cleaning machines.

[0071] In some embodiments, a first-level pre-inspection is performed using environmental data collected by a pre-inspection sensing module, that is, the environmental matching degree between the main edge cleaning site and the slave edge cleaning site is judged based on the site pre-inspection data and site actual data collected by the pre-inspection sensing module, and whether to execute the reference cleaning path is determined based on the environmental matching degree.

[0072] The method also includes: when at the main edge cleaning site, enabling the pre-inspection sensing module, acquiring the environmental data collected by the pre-inspection sensing module and storing it as site pre-inspection data; when at the secondary edge cleaning site, enabling the pre-inspection sensing module, and using the environmental data collected by the pre-inspection sensing module as site actual data; comparing the data error between the site pre-inspection data and the site actual data, and if the data error is less than a preset environmental error threshold, controlling the photovoltaic robot to execute the reference cleaning path.

[0073] The pre-inspection sensing module may be a visual sensing module, a laser radar, an ultrasonic sensor, an infrared sensor, etc., and may be flexibly determined according to the sensing module equipped by the photovoltaic robot, and a sensing module that is easier to judge the similarity of the environmental data of the adjacent position may be selected. This embodiment is described by taking the visual sensing module as an example.

[0074] Specifically, at the main edge cleaning site, the pre-inspection perception module is enabled, which is responsible for collecting the site pre-inspection data of the boundary cleaning site, such as photovoltaic panel characteristics, surrounding environment characteristics, photovoltaic panel boundary characteristics, etc. The environmental data collected by the pre-inspection perception module is stored as site pre-inspection data for subsequent path planning and verification. When at the slave edge cleaning site, the pre-inspection perception module is enabled again to collect the actual site data of the current environment. The site pre-inspection data is compared with the actual site data, and the data error between the two is calculated. If the data error is less than the preset environmental error threshold, it means that the environment of the main edge cleaning site and the slave edge cleaning site match, the reference cleaning path can be trusted, and the photovoltaic robot is controlled to perform the cleaning task according to the reference cleaning path. Among them, the preset environmental error threshold can be flexibly set according to actual needs, and is not limited here.

[0075] In some embodiments, when the pre-check fails, deep edge detection is performed at the cleaning station, and the reliability of the reference cleaning path is verified again based on the environmental data and the cleaning path during the deep edge detection process.

[0076] Specifically, after comparing the data error between the site pre-inspection data and the site actual data, it also includes: if the data error is greater than the preset environmental error threshold, updating the edge detection cleaning site to the verification edge detection site, replacing other perception modules as verification perception modules, obtaining the environmental data collected by the verification perception module and storing them as site verification data; performing deep edge detection on the boundary to be cleaned, generating a verification cleaning path for the verification edge detection site; comparing the verification cleaning path with the reference cleaning path to obtain a path error; if the path error is less than the preset path error threshold, retaining the reference cleaning path, updating the verification perception module to the pre-inspection perception module, and updating the site verification data to the site pre-inspection data. Among them, the preset path error threshold can be flexibly set according to actual needs and is not limited here.

[0077] It should be understood that replacing other perception modules as verification perception modules can avoid misjudgment due to pre-inspection perception module failure and eliminate data interference from the pre-inspection perception module. When it is confirmed that the path error is less than the preset path error threshold, it is confirmed that the pre-inspection perception module has detected an error. Replacing the site pre-inspection data and pre-inspection perception module avoids security issues caused by data uniqueness or accuracy limitations and improves the accuracy of reliability testing.

[0078] In some embodiments, after the simple edge detection is completed, a secondary re-inspection is performed through the actual cleaning path from the edge detection cleaning station, that is, the similarity between the actual cleaning path and the reference cleaning path is compared, and whether the reference cleaning path is retained is determined according to the similarity.

[0079] The method includes: after controlling the photovoltaic robot to stop moving, obtaining the actual cleaning path of the photovoltaic robot; comparing the actual cleaning path with the reference cleaning path to obtain a path error, and if the path error is less than a preset path error threshold, retaining the reference cleaning path. The preset path error threshold can be flexibly set according to actual needs and is not limited here.

[0080] It should be understood that before the reference cleaning path is completed, the perception module recognizes the boundary in advance and controls the photovoltaic robot to stop moving, resulting in errors. There are two specific reasons. One is that the reference cleaning path is accurate, but due to different sensor recognition sensitivities or unstable data, the boundary is recognized in advance and the photovoltaic robot is controlled to stop moving; the other is that the photovoltaic panel is damaged, resulting in an inaccurate reference cleaning path.

[0081] Therefore, by comparing the actual cleaning path with the reference cleaning path, the path error is calculated, and the above two causes are distinguished by the size of the path error to confirm whether the reference cleaning path can be trusted.

[0082] In some embodiments, the primary pre-check scheme and the secondary re-check scheme may be enabled simultaneously, or only one of the checks may be enabled to confirm whether the reference cleaning path is trustworthy.

[0083] In some embodiments, the method further includes: if the path error is greater than the preset path error threshold, deleting the reference cleaning path, and updating all boundary cleaning stations within the preset spacing area to independent edge detection stations; when reaching the independent edge detection station, enabling all the perception modules to perform deep edge detection on the boundary to be cleaned, and generating a cleaning path for each independent edge detection station.

[0084] The preset spacing area is the area extending from the current boundary cleaning station along the same boundary to the uncleaned area by the preset untrusted distance, and the preset untrusted distance can be flexibly set according to actual conditions. For example, the preset spacing area can be set to 1m, that is, the area extending from the boundary cleaning station along the same boundary to the uncleaned area by 1m.

[0085] It should be understood that when the reliability of the reference cleaning path is found to be in doubt in the first-level pre-inspection plan and the second-level re-inspection plan, the differentiated perception module scheduling strategy will be suspended within the preset spacing area, and all perception modules will be used for precise boundary exploration, with safety as the priority, to ensure cleaning coverage and improve the reliability and practicality of the photovoltaic robot.

[0086] Exemplarily, the preset spacing area is the length area or width area of ​​a single photovoltaic panel laid out on the photovoltaic panel map. Since the shape of the photovoltaic panel is a regular rectangle, according to the array arrangement of the photovoltaic panels, the long side or wide side of the rectangle is determined to be the preset spacing area. When the reliability of the reference cleaning path is found to be questionable, the differentiated perception module scheduling strategy is suspended within a photovoltaic panel area, and the size of the preset non-trust distance is reasonably limited under the premise of ensuring safety.

[0087] In some embodiments, the present invention provides a path planning method for a highly adaptable photovoltaic robot that improves cleaning efficiency while ensuring cleaning quality. Various photovoltaic robots with different cleaning ranges and cleaning capabilities are adapted to flexibly determine the spacing between adjacent paths, and customized internal cleaning paths that meet the expected cleaning effects are generated to ensure that the planning of the cleaning path can be scalably and accurately adapted to the needs of different cleaning tasks. Furthermore, during the cleaning of the photovoltaic panel, the execution of the cleaning task is monitored in real time from two dimensions: path deviation and photovoltaic panel cleaning effect, and different processing modes are adopted for different degrees of path execution errors.

[0088] In some embodiments, the photovoltaic robot includes at least two perception modules, and the perception module includes a visual perception module. Figure 6 , Figure 6 is a schematic flow chart of a path planning method for a photovoltaic robot provided by an embodiment of the present invention, such as Figure 6 As shown, it includes steps S201 to S205.

[0089] S201 obtains a machine cleaning performance index, and sets an adjacent path spacing between adjacent sub-cleaning paths according to the machine cleaning performance index.

[0090] The cleaning performance index refers to one or more parameters for evaluating the cleaning efficiency of the photovoltaic robot, and then the adjacent path spacing between adjacent sub-cleaning paths is comprehensively determined based on one or more parameters. The smaller the adjacent path spacing, the more times the photovoltaic robot repeatedly cleans the same area, and the corresponding cleaning effect is better. Among them, the cleaning efficiency is reflected in the cleaning power, cleaning operation endurance and other aspects.

[0091] In some embodiments, the cleaning performance index includes a cleaning force index and a cleaning coverage index. S201 further includes: determining a minimum number of repeated cleaning times according to the cleaning force index and an expected cleaning standard; and determining the adjacent path spacing between adjacent sub-cleaning paths according to the cleaning coverage index and the minimum number of repeated cleaning times.

[0092] Among them, the cleaning power index refers to the degree of dirt removal that the photovoltaic robot can achieve in a single cleaning. It is one of the parameters for evaluating the cleaning ability of the photovoltaic robot. The higher the cleaning power, the fewer cleaning times can be used to achieve the same cleaning effect.

[0093] Among them, the cleaning coverage index refers to the area that the cleaning device of the photovoltaic robot can cover. A larger coverage area means that cleaning work over a large area can be completed faster.

[0094] Among them, the expected cleaning standards may include the degree of cleaning, the area that can be cleaned on a single charge, and other requirements that users put forward for the entire process of photovoltaic panel cleaning from different angles.

[0095] For example, based on the cleaning force index, the minimum number of repeated cleanings required to achieve the expected cleaning standard for the same area can be calculated. According to the cleaning coverage index and the minimum number of repeated cleanings, the distance that should be maintained between each sub-cleaning path, that is, the spacing between adjacent paths, can be calculated so that after the photovoltaic panel is cleaned the minimum number of repeated cleanings, the cleanliness of the photovoltaic panel surface can reach the expected cleaning standard, thereby improving the overall cleaning quality.

[0096] For example, if the cleaning force required for complete and thorough cleaning is 100%, the cleaning force index of the photovoltaic robot is 40%, and the expected cleaning standard cleaning degree is 80%, then the minimum number of repeated cleaning times is 2 times. When the cleaning coverage index is 0.4m, the adjacent path spacing can be set to 0.2m so that each area to be cleaned is cleaned twice. It should be understood that the area size where the cleaning coverage index exceeds the adjacent path spacing is the area size of the repeated cleaning area.

[0097] In some embodiments, the cleaning performance indicators include a battery life indicator. Based on the battery life indicator and the photovoltaic panel map of the photovoltaic robot, the power required for the photovoltaic robot to clean the current photovoltaic panel map can be calculated. According to the number of charging times corresponding to the expected cleaning standard, the minimum path spacing between each sub-cleaning path is calculated; when the minimum path spacing is greater than the cleaning coverage indicator, an insufficient cleaning signal is generated to remind the user.

[0098] Furthermore, when the minimum path spacing is less than the cleaning coverage index, the adjacent path spacing is determined based on the cleaning coverage index, the cleaning force index and the minimum path spacing, and the final adjacent path spacing is greater than the minimum path spacing to take into account both energy consumption and cleaning quality.

[0099] It should be understood that formulating different cleaning paths according to cleaning performance indicators can effectively adapt to various types of photovoltaic robots with different cleaning ranges and cleaning capabilities. For example, as the number of uses increases, the photovoltaic robot may experience different degrees of loss. For example, if the cleaning ability decreases, repeated cleaning is required to ensure the cleaning effect, and a smaller adjacent path spacing can be set accordingly; for example, if the battery life decreases, it is necessary to avoid repeated cleaning to reduce unnecessary energy consumption, and a larger adjacent path spacing can be set accordingly. In this way, the cleaning path of the photovoltaic robot is planned with high adaptability, while taking into account both cleaning quality and endurance.

[0100] S202 obtains a photovoltaic panel map, generates an internal cleaning path based on the photovoltaic panel map and the adjacent path spacing, and discretizes the internal cleaning path into a plurality of path points, wherein the parameters of the path points include an expected posture.

[0101] The internal cleaning path includes multiple sub-cleaning paths, each of which covers the internal area of ​​the photovoltaic panel map that is relatively far from the boundary. In order to achieve precise posture control, the internal cleaning path is discretized into several path points. For example, the distance between path points is about 5 mm. Each path point is assigned reasonable path point parameters such as position, posture, speed, angular velocity, etc. according to the cleaning task, which are used to determine in real time whether the photovoltaic robot deviates from the path and the degree of path deviation, so as to ensure the stability and efficiency of the robot during movement.

[0102] Exemplarily, based on the path planning algorithm and the adjacent path spacing, an internal cleaning path is generated for the internal area of ​​the photovoltaic panel map, and the path spacing between each sub-cleaning path of the internal cleaning path is the adjacent path spacing, so as to generate a customized internal cleaning path that meets the expected cleaning effect, and ensure that the planning of the cleaning path can be scalably and accurately adapted to different cleaning tasks and different photovoltaic robots. For details, please refer to the description in other embodiments of the present invention, which will not be repeated here.

[0103] It should be understood that the path points and cleaning stations in the embodiments of the present invention are different. In order to fully clean the photovoltaic panel array, the internal cleaning path needs to fully cover the internal area. Therefore, there are multiple turning or speed change positions in the cleaning path. Cleaning stations are set at these positions to control the photovoltaic robot to change speed or turn in time, so as to accurately execute the internal cleaning path. In addition, in order to ensure the smooth movement of the photovoltaic robot, the speed change or turn of the photovoltaic robot is gradually completed during the movement. Correspondingly, the parameters of the path points around the cleaning station are generated according to the speed or direction to be adjusted to ensure that the photovoltaic robot can smoothly and accurately complete the speed change or turn.

[0104] For example, the photovoltaic robot changes speed at cleaning station A, accelerating from 0.1m / s to 0.4m / s. The corresponding cleaning path within the preset speed change distance before cleaning station A or the preset speed change distance after cleaning station A is the acceleration path. The acceleration path is discretized into path point q and path point p. The photovoltaic robot passes through path point q and path point p in turn. Correspondingly, the speed parameter value of path point q is 0.2m / s, and the speed parameter value of path point p is 0.3m / s. Therefore, when the photovoltaic robot reaches cleaning station A after traveling the preset speed change distance, it accelerates to 0.4m / s; or after traveling the preset speed change distance from cleaning station A, the photovoltaic robot accelerates to 0.4m / s.

[0105] S203: setting an acceptance path point on each sub-cleaning path of the internal cleaning path, wherein parameters of the acceptance path point include an expected cleaning area.

[0106] The acceptance path point is a location point where the cleaning quality is checked during the cleaning process. The expected cleaning area is the area that is expected to have been cleaned in the image collected at the current acceptance path point.

[0107] For example, in order to improve the usability of the image, the relative angle between the acceptance path points is less than the preset angle value, ensuring that the angle difference between the acceptance path points is controllable, so that the continuity and coverage of the image between adjacent sub-cleaning paths can be ensured, ensuring that each part of the photovoltaic panel can be effectively cleaned and inspected, thereby ensuring that the cleaning quality of the photovoltaic panel meets the expected standards and improving the reliability and effectiveness of the entire cleaning process. Figure 7 , Figure 7 is a schematic diagram of an acceptance path point provided by an embodiment of the present invention, such as Figure 7 As shown, the relative angles between the acceptance path points are close to zero.

[0108] S204 controls the photovoltaic robot to execute the internal cleaning path; monitors in real time the posture error between the actual posture collected by the perception module and the expected posture; at each acceptance path point, obtains the actual cleaning area based on the visual perception module, and determines the cleaning performance error according to the actual cleaning area and the expected cleaning area.

[0109] Specifically, the photovoltaic robot is controlled to operate along a predetermined internal cleaning path, while the actual posture collected by the perception module is monitored in real time. By comparing the actual posture data with the expected posture data, the posture error can be measured. Furthermore, the cleaning results of the current photovoltaic robot are evaluated at the path acceptance point through images to obtain the actual cleaning area. Based on the actual cleaning area and the expected cleaning area, it can be determined whether there is a cleaning performance error in the cleaning performance index of the photovoltaic robot.

[0110] That is to say, during the process of the photovoltaic robot executing the internal cleaning path, through real-time analysis of posture errors and fixed-point monitoring of cleaning effectiveness, the execution of the cleaning task is monitored in real time from two dimensions: path deviation and photovoltaic panel cleaning effect. The cleaning strategy of the photovoltaic robot is adjusted appropriately in practice to ensure the cleaning quality and travel safety, so as to ensure that the cleaning work of the photovoltaic panels achieves the expected results.

[0111] In some embodiments, when the photovoltaic robot reaches the acceptance path point, the image acceptance data collected by the visual perception module is obtained; based on a preset recognition model, the final cleaning area image in the image acceptance data is identified; and based on a preset mapping algorithm, the actual cleaning area corresponding to the final cleaning area image is calculated.

[0112] Specifically, an acceptance path point is set on each sub-cleaning path. When the photovoltaic robot reaches the acceptance path point, the image acceptance data collected by the visual perception module is obtained and stored for subsequent cleaning effect evaluation and quality control. The final cleaning area image of the cleaning path that has completed all cleaning paths in the image acceptance data is identified according to the trained preset recognition model, and then the actual cleaning area corresponding to the final cleaning area image is calculated based on the preset mapping algorithm.

[0113] Among them, the preset mapping algorithm is generated based on the proportional relationship between the image area and the actual area.

[0114] Among them, the preset recognition model is a cleanliness recognition model based on machine learning or deep learning technology, which can analyze and classify images. Exemplarily, the training process of the preset recognition model uses three different types of photovoltaic panel pictures: fully cleaned, partially cleaned, and uncleaned. Different cleaning powers are distinguished by parameters such as photovoltaic panel cleanliness, dust adhesion, and transmittance. The corresponding data such as brightness, texture features, and grayscale values ​​of different parameters in the image data are also different. These pictures are used as training data to help the model learn how to identify and distinguish photovoltaic panels with different cleaning powers, and then the cleanliness recognition model obtained by training can receive image acceptance data as input and distinguish the fully cleaned, partially cleaned, or uncleaned area images in the current image acceptance data.

[0115] For example, when the dust adhesion is low, the transmittance will increase, and the corresponding cleanliness is high. Such photovoltaic panels can reflect more sunlight, making the corresponding area in the image appear brighter. On the contrary, when the dust adhesion is high, the transmittance will decrease, and the corresponding cleanliness is low. The dirt or dust on the surface of the photovoltaic panel will absorb more light, causing the corresponding area in the image to appear darker. The shadows and irregular shapes of the dust may also cause dark spots or uneven textures in the image.

[0116] It should be understood that in order to adapt to various types of photovoltaic robots with different cleaning ranges and cleaning capabilities, the photovoltaic robot may need to repeatedly clean the same area. The final cleaning area is the area that has been repeatedly cleaned. For example, each area to be cleaned needs to be cleaned twice, then the final cleaning area is the area that has been cleaned twice.

[0117] In some embodiments, a photovoltaic robot path planning method provided by an embodiment of the present invention further includes a slow response step S206: when the posture error is less than the posture tolerance threshold and the cleaning performance error is less than the cleanliness tolerance threshold, controlling the photovoltaic robot to continue executing the internal cleaning path.

[0118] Among them, the posture tolerance threshold can be determined based on the preset distance value between the internal area and the boundary area. For example, when the preset distance value is 50 cm, the risk of falling of the photovoltaic robot increases once it deviates by 50 cm. Therefore, in order to ensure the safe movement of the photovoltaic robot, the posture tolerance threshold is set to 50 cm.

[0119] In addition, the photovoltaic panel is composed of multiple electronic boards. The installation gap between the electronic boards will form multiple lines. The visual perception module will collect limited images and fit the lines in the image into multiple white lines. Furthermore, the limited image includes at least one baseline and two reference lines that are updated in real time. The posture changes of the photovoltaic robot are determined according to the changes of these lines in the field of view. When the position offset distance of the photovoltaic robot is too large, it is difficult to return to the expected position through the visual perception module. At this time, the posture tolerance threshold can also be set based on the posture calibration capability of the perception module.

[0120] Among them, the cleanliness tolerance threshold can be flexibly set according to actual needs.

[0121] It should be noted that when setting the adjacent path spacing between adjacent sub-cleaning paths according to the machine cleaning performance index, it is often difficult to properly match the cleaning force index with the expected cleaning standard through the number of repeated cleanings. For example, the cleaning force index is 50%, and the expected cleaning standard is 80% of the cleaning degree. At this time, in order to ensure the cleaning quality, the minimum number of repeated cleanings is 2 times, and the actual cleaning degree can theoretically reach 100%. Therefore, the cleaning degree of the fully cleaned area has exceeded the expected cleaning standard. From the overall point of view, even if there is a small area that is not fully cleaned, the power generation efficiency of the photovoltaic panel has reached a significant improvement in the expected cleaning standard, and the smaller cleaning omissions caused by this offset can be tolerated.

[0122] Furthermore, the redundancy rate is determined according to the minimum number of repeated cleanings, the cleaning power index, and the expected cleaning standard. The cleanliness tolerance threshold can also be flexibly set according to the redundancy rate. For example, the cleanliness tolerance threshold can be set to 0.5 square meters at this time.

[0123] It should be understood that when cleaning the inner area of ​​the photovoltaic panel array, the movement of the photovoltaic robot in this area is relatively safe, and some areas will be cleaned repeatedly to ensure the cleaning quality, thus allowing the photovoltaic robot to enjoy a certain degree of freedom in the execution of the path.

[0124] That is to say, even if there is a path execution error within this degree of freedom, the repeated cleaning strategy can ensure that the cleaning quality is maintained stably, and the regular paving of the internal area and the low risk of falling can ensure the safety of the photovoltaic robot. The embodiment of the present invention uses the posture tolerance threshold and the clean tolerance threshold to monitor in real time whether the current path execution error is within this degree of freedom from two dimensions: path deviation and photovoltaic panel cleaning effect. The tolerable path execution error corresponding to this degree of freedom is not adjusted, avoiding frequent adjustments to the cleaning path, reducing unnecessary repetitions of the cleaning path, improving the photovoltaic robot's execution speed and cleaning efficiency of the cleaning path, saving energy and time, and at the same time, the cleaning quality and travel safety are still reliably guaranteed.

[0125] In some embodiments, a path planning method for a photovoltaic robot provided by an embodiment of the present invention includes a quick response step S205: when the posture error is greater than the posture tolerance threshold or the cleaning performance error is greater than the cleanliness tolerance threshold, a target internal cleaning path is generated based on the posture error and / or the cleaning performance error, and the photovoltaic robot is controlled to execute the target internal cleaning path.

[0126] Specifically, if the posture error exceeds the preset posture tolerance threshold, or the cleaning performance error is greater than the set cleanliness tolerance threshold, it is considered that the actual position of the robot deviates too much from the expected path or the cleaning quality is difficult to meet the expected standard. A new target internal cleaning path is generated by comprehensively considering the posture error and the cleaning performance error to correct the path execution error. The photovoltaic robot will continue to perform the cleaning task according to this newly generated target internal cleaning path.

[0127] It should be understood that when the cleaning quality and travel safety cannot be effectively guaranteed, the path execution exceeds the allowed degrees of freedom, and the position posture and path spacing are adjusted in time according to real-time data to ensure the cleaning quality and safety and improve the effectiveness and flexibility of the cleaning task.

[0128] In some embodiments, the target internal cleaning path includes a posture correction path and an optimized cleaning path, and S205 includes: generating a posture correction path of the photovoltaic robot based on the posture error; and / or updating a machine cleaning performance index based on the cleaning performance error; updating an adjacent path spacing between adjacent sub-cleaning paths according to the updated machine cleaning performance index; generating the optimized cleaning path based on the updated adjacent path spacing, and controlling the photovoltaic robot to execute the optimized cleaning path.

[0129] Among them, the posture correction path is a path for correcting the posture error, and the optimized cleaning path is a path obtained by adjusting the adjacent path spacing of the unexecuted internal cleaning path.

[0130] Specifically, when the posture error is greater than the posture tolerance threshold, a posture correction path is generated based on the posture error of the photovoltaic robot and the path planning algorithm to correct the photovoltaic robot to the expected posture and ensure the safety of the photovoltaic robot. When the cleaning performance error is greater than the cleanliness tolerance threshold, the adjacent path spacing that is more suitable for the current application scenario is determined, and an optimized cleaning path is generated based on the path planning algorithm and the updated adjacent path spacing to ensure that the cleaning quality meets the expected standards.

[0131] In some embodiments, when the posture error is greater than the posture tolerance threshold and the cleaning performance error is less than the clean tolerance threshold, or when the cleaning performance error is greater than the clean tolerance threshold and the posture error is less than the posture tolerance threshold, a posture correction path and an optimized cleaning path can be generated respectively, and the posture error and the cleaning performance error can be eliminated through a one-time path adjustment, thereby avoiding subsequent frequent adjustments to the cleaning path, reducing unnecessary repetitions of the cleaning path, and improving cleaning efficiency while ensuring cleaning quality.

[0132] It should be understood that the photovoltaic array cleaning method provided by the embodiment of the present invention can dynamically adapt to changes in the cleaning performance of the photovoltaic robot. For example, as the photovoltaic robot performs the cleaning task, its cleaning performance index may gradually decrease, and at this time, the adjacent path spacing needs to be reduced; for another example, when the obtained machine cleaning performance index is inaccurate, the cleaning performance error is greater than the clean tolerance threshold, which may be that the machine cleaning performance index may be too high or too low, and the adjacent path spacing can be increased or decreased accordingly to reduce repeated cleaning or missed cleaning.

[0133] In some embodiments, the S205 also includes: controlling the photovoltaic robot to move according to the posture correction path; obtaining a corrected posture based on the perception module, and if the posture error between the corrected posture and the expected posture is less than a preset correction error threshold, the posture correction is successful; controlling the photovoltaic robot to execute the optimized cleaning path.

[0134] It should be understood that the posture correction path is generated and executed based on the posture error, and the photovoltaic robot is judged whether it is corrected to the expected position in the photovoltaic panel array by matching the corrected posture with the expected posture, so as to ensure the accuracy of the posture correction. Furthermore, if the posture error between the corrected posture and the expected posture is greater than the preset correction error threshold, other perception modules are called to re-correct the posture until the posture correction is successful.

[0135] In some embodiments, the target internal cleaning path includes a compensatory cleaning path, and S205 also includes: when the cleaning performance error is lower than the clean tolerance threshold, generating and executing a compensatory cleaning path for the area within the image range captured by the acceptance path point, so as to perform compensatory cleaning on the area that is not fully cleaned, thereby ensuring that the cleaning work meets the expected standards and improving the overall cleaning quality.

[0136] In some embodiments, a minimum spacing value of adjacent path spacing between adjacent sub-cleaning paths is set. When the updated adjacent path spacing is less than the fault spacing value, the cleaning performance index of the photovoltaic robot is too low. Continuing cleaning may result in ineffective cleaning that wastes electricity. The internal cleaning path can be stopped and the user can be reminded.

[0137] In some embodiments, if the posture error monitored at high frequency in the path is greater than the posture tolerance threshold or the cleaning performance error is greater than the cleanliness tolerance threshold, considering special circumstances such as map distortion and machine failure, the internal cleaning path is stopped and the user is reminded.

[0138] In some embodiments, the method also includes: obtaining the fluctuation frequency of the actual posture; when the fluctuation frequency is not within a preset frequency range, identifying whether there is a high-risk area on the cleaning path based on the environmental data collected by the perception module, and if there is no high-risk area on the cleaning path, controlling the photovoltaic robot to continue to execute the internal cleaning path; if there is a high-risk area on the cleaning path, controlling the photovoltaic robot to bypass the high-risk area and continue to execute the internal cleaning path.

[0139] When the photovoltaic robot is moving smoothly and normally, the posture change should be within the preset frequency range. The fluctuation frequency of the actual posture should be evaluated, and the posture mutations caused by high-risk areas (such as damaged areas and slippery areas) should be identified based on the environmental data collected by the perception module. The cleaning path should be flexibly adjusted to bypass the high-risk areas, and additional monitoring should be performed on abnormal fluctuations in data within the tolerance range of path execution errors to adapt to different environmental conditions and photovoltaic panel conditions, thereby improving the adaptability, safety and reliability of the photovoltaic robot.

[0140] Through this sophisticated monitoring and analysis, manual intervention is reduced, operation and maintenance costs are lowered, energy waste can be minimized and the risk of incomplete cleaning can be avoided, thereby improving the economy and sustainability of photovoltaic power generation systems.

[0141] An embodiment of the present invention further provides a photovoltaic robot, the photovoltaic robot comprising: A moving device for moving between photovoltaic panels; A detection device is installed on the mobile device, and the detection device includes at least two perception modules, and the perception modules are used to collect motion data or environmental data of the photovoltaic robot; wherein the perception module may include a visual perception module.

[0142] A cleaning device, installed on the mobile device, for cleaning the photovoltaic panel; A control system is connected to the mobile device, the detection device and the cleaning device, and is used to implement the steps of the photovoltaic array cleaning method and / or the photovoltaic robot path planning method provided in any embodiment of the present invention. For details, please refer to the description of the photovoltaic array cleaning method and / or the photovoltaic robot path planning method in other embodiments of the present invention, which will not be repeated here.

[0143] See also Figure 8 , Figure 8 1 is a schematic block diagram of a computer device provided by an embodiment of the present invention. The computer device may be a server.

[0144] The computer device includes a processor, a memory and a network interface connected via a system bus, wherein the memory may include a non-volatile storage medium and an internal memory.

[0145] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the photovoltaic array cleaning methods and / or photovoltaic robot path planning methods.

[0146] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0147] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any one of the photovoltaic array cleaning methods and / or photovoltaic robot path planning methods.

[0148] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0149] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0150] In one embodiment, the photovoltaic robot includes at least two sensing modules, and the processor is used to run a computer program stored in a memory to implement the following steps: S101: obtaining a photovoltaic panel map, and generating an internal cleaning path based on the photovoltaic panel map, wherein the internal cleaning path passes through a plurality of cleaning stations, and the cleaning stations include boundary cleaning stations with a preset distance value from any boundary to be cleaned on the photovoltaic panel map; S102 controls the photovoltaic robot to execute the internal cleaning path, and when reaching the boundary cleaning station, detects whether there is a main edge detection cleaning station of the same boundary to be cleaned in the trust interval area; S103: if there is no main edge detection cleaning station for the same boundary to be cleaned, setting the boundary cleaning station as the main edge detection cleaning station, enabling all the perception modules to perform deep edge detection on the boundary to be cleaned, and generating a reference cleaning path for the main edge detection cleaning station; S104: if there is a main edge detection cleaning station for the same boundary to be cleaned, the boundary cleaning station is set as a secondary edge detection cleaning station, the photovoltaic robot is controlled to obtain and execute the reference cleaning path of the main edge detection cleaning station, and part of the perception module is enabled to perform simple edge detection on the boundary to be cleaned.

[0151] In some embodiments, the trust distance area is a length area or a width area of ​​a single photovoltaic panel laid out on the photovoltaic panel map.

[0152] In some embodiments, when implementing the step S103, the processor is also used to implement: controlling the photovoltaic robot to move toward the boundary to be cleaned based on a first speed and a preset edge detection value, wherein the preset edge detection value is greater than the preset distance value; enabling all the perception modules to collect multiple environmental data in real time, and performing boundary identification based on the multiple environmental data, and when the boundary is identified, controlling the photovoltaic robot to stop moving; obtaining the moving path of the photovoltaic robot as a reference cleaning path for the main edge detection cleaning station.

[0153] In some embodiments, when implementing the step S104, the processor is also used to implement: controlling the photovoltaic robot to execute the reference cleaning path based on the second speed, enabling any perception module to collect environmental data in real time, and performing boundary identification based on the environmental data; when the boundary is identified and / or the reference cleaning path is completed, controlling the photovoltaic robot to stop moving.

[0154] In some embodiments, the method further includes: at the main edge cleaning site, enabling the pre-inspection sensing module, acquiring the environmental data collected by the pre-inspection sensing module and storing it as site pre-inspection data; at the slave edge cleaning site, enabling the pre-inspection sensing module, and using the environmental data collected by the pre-inspection sensing module as site actual data; comparing the data error between the site pre-inspection data and the site actual data, if the data error is less than a preset environmental error threshold, controlling the photovoltaic robot to execute the reference cleaning path; and / or, after controlling the photovoltaic robot to stop moving, acquiring the actual cleaning path of the photovoltaic robot; comparing the actual cleaning path with the reference cleaning path to obtain a path error, if the path error is less than a preset path error threshold, retaining the reference cleaning path.

[0155] In some embodiments, after implementing the comparison of the data error between the site pre-inspection data and the site actual data, the processor is also used to implement: if the data error is greater than the preset environmental error threshold, updating the edge detection cleaning site to a verification edge detection site, replacing other perception modules as verification perception modules, obtaining environmental data collected by the verification perception module and storing them as site verification data; performing deep edge detection on the boundary to be cleaned, and generating a verification cleaning path for the verification edge detection site; comparing the verification cleaning path with the reference cleaning path to obtain a path error; if the path error is less than the preset path error threshold, retaining the reference cleaning path, updating the verification perception module to a pre-inspection perception module, and updating the site verification data to the site pre-inspection data.

[0156] In some embodiments, the processor is also used to implement: if the path error is greater than the preset path error threshold, the reference cleaning path is deleted, and all boundary cleaning stations within the preset spacing area are updated to independent edge detection stations; when reaching the independent edge detection station, all the perception modules are enabled to perform deep edge detection on the boundary to be cleaned, and a cleaning path for each independent edge detection station is generated.

[0157] In one embodiment, the photovoltaic robot includes at least two perception modules, the perception module may include a visual perception module, and the processor is used to run a computer program stored in the memory to implement the following steps: S201: obtaining a machine cleaning performance index, and setting an adjacent path spacing between adjacent sub-cleaning paths according to the machine cleaning performance index; S202: obtaining a photovoltaic panel map, generating an internal cleaning path based on the photovoltaic panel map and the adjacent path spacing, and discretizing the internal cleaning path into a plurality of path points, wherein the parameters of the path points include an expected posture; S203 setting an acceptance path point on each sub-cleaning path of the internal cleaning path, wherein parameters of the acceptance path point include an expected cleaning area; S204 controls the photovoltaic robot to execute the internal cleaning path; monitors in real time the posture error between the actual posture acquired by the perception module and the expected posture; obtains the actual cleaning area based on the visual perception module at each acceptance path point, and determines the cleaning performance error according to the actual cleaning area and the expected cleaning area; S205: When the posture error is greater than the posture tolerance threshold or the cleaning performance error is greater than the cleaning tolerance threshold, a target internal cleaning path is generated based on the posture error and / or the cleaning performance error, and the photovoltaic robot is controlled to execute the target internal cleaning path.

[0158] In some embodiments, the processor is further configured to: when the posture error is less than a posture tolerance threshold and the cleaning performance error is less than a cleanliness tolerance threshold, control the photovoltaic robot to continue executing the internal cleaning path.

[0159] In some embodiments, the target internal cleaning path includes a posture correction path and an optimized cleaning path. When implementing the step S205, the processor is also used to implement: generating a posture correction path for the photovoltaic robot based on the posture error; and / or updating a machine cleaning performance index based on the cleaning performance error; updating an adjacent path spacing between adjacent sub-cleaning paths according to the updated machine cleaning performance index; generating the optimized cleaning path based on the updated adjacent path spacing, and controlling the photovoltaic robot to execute the optimized cleaning path.

[0160] In some embodiments, when implementing the step S205, the processor is also used to implement: controlling the photovoltaic robot to move according to the posture correction path; obtaining the corrected posture based on the perception module, if the posture error between the corrected posture and the expected posture is less than a preset correction error threshold, the posture correction is successful; controlling the photovoltaic robot to execute the optimized cleaning path.

[0161] In some embodiments, when implementing the step S204, the processor is also used to implement: when the photovoltaic robot reaches the acceptance path point, obtaining the image acceptance data collected by the visual perception module; based on a preset recognition model, identifying the final cleaning area image in the image acceptance data; and calculating the actual cleaning area corresponding to the final cleaning area image based on a preset mapping algorithm.

[0162] In some embodiments, the cleaning performance index includes a cleaning force index and a cleaning coverage index. When implementing the step S201, the processor is also used to implement: determining the minimum number of repeated cleaning times based on the cleaning force index and the expected cleaning standard; determining the adjacent path spacing between adjacent sub-cleaning paths based on the cleaning coverage index and the minimum number of repeated cleaning times.

[0163] In some embodiments, the processor is also used to implement: obtaining the fluctuation frequency of the actual posture; when the fluctuation frequency is not within a preset frequency range, identifying whether there is a high-risk area on the cleaning path based on the environmental data collected by the perception module, and if there is no high-risk area on the cleaning path, controlling the photovoltaic robot to continue to execute the internal cleaning path; if there is a high-risk area on the cleaning path, controlling the photovoltaic robot to bypass the high-risk area and continue to execute the internal cleaning path.

[0164] A computer-readable storage medium is also provided in an embodiment of the present invention, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and the processor executes the program instructions to implement any one of the photovoltaic array cleaning methods and / or photovoltaic robot path planning methods provided in the embodiments of the present invention.

[0165] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the computer device.

[0166] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A path planning method for a photovoltaic robot, characterized in that: The photovoltaic robot includes at least two perception modules, wherein the perception module includes a visual perception module, and the method includes: S201: obtaining a machine cleaning performance index, and setting an adjacent path spacing between adjacent sub-cleaning paths according to the machine cleaning performance index; S202: obtaining a photovoltaic panel map, generating an internal cleaning path based on the photovoltaic panel map and the adjacent path spacing, and discretizing the internal cleaning path into a plurality of path points, wherein the parameters of the path points include an expected posture; S203 setting an acceptance path point on each sub-cleaning path of the internal cleaning path, wherein parameters of the acceptance path point include an expected cleaning area; S204 controls the photovoltaic robot to execute the internal cleaning path; monitors in real time the posture error between the actual posture acquired by the perception module and the expected posture; obtains the actual cleaning area based on the visual perception module at each acceptance path point, and determines the cleaning performance error according to the actual cleaning area and the expected cleaning area; S205: When the posture error is greater than the posture tolerance threshold or the cleaning performance error is greater than the cleaning tolerance threshold, a target internal cleaning path is generated based on the posture error and / or the cleaning performance error, and the photovoltaic robot is controlled to execute the target internal cleaning path.

2. The method according to claim 1, characterized in that: The method further comprises: When the posture error is less than the posture tolerance threshold and the cleaning performance error is less than the cleanliness tolerance threshold, the photovoltaic robot is controlled to continue executing the internal cleaning path.

3. The method according to claim 1, characterized in that The target internal cleaning path includes a posture correction path and an optimized cleaning path, and S205 includes: generating a posture correction path of the photovoltaic robot based on the posture error; and / or, updating a machine cleaning performance indicator based on the cleaning performance error; updating the adjacent path spacing between adjacent sub-cleaning paths according to the updated machine cleaning performance indicator; The optimized cleaning path is generated based on the updated adjacent path spacing, and the photovoltaic robot is controlled to execute the optimized cleaning path.

4. The method according to claim 3, characterized in that The S205 further includes: Controlling the photovoltaic robot to move according to the posture correction path; Acquire a corrected posture based on the perception module, and if a posture error between the corrected posture and the expected posture is less than a preset correction error threshold, the posture correction is successful; The photovoltaic robot is controlled to execute the optimized cleaning path.

5. The method according to claim 1, characterized in that The S204 further includes: When the photovoltaic robot reaches the acceptance path point, acquiring image acceptance data collected by the visual perception module; Based on a preset recognition model, identifying a final clean area image in the image acceptance data; The actual cleaning area corresponding to the final cleaning region image is calculated based on a preset mapping algorithm.

6. The method according to any one of claims 1 to 5, characterized in that: The cleaning performance index includes a cleaning power index and a cleaning coverage index. S201 further includes: Determine the minimum number of repeated cleanings based on the cleaning power index and the expected cleaning standard; The adjacent path spacing between adjacent sub-cleaning paths is determined according to the cleaning coverage index and the minimum number of repeated cleaning times.

7. The method according to claim 1, characterized in that The method further comprises: Obtaining a fluctuation frequency of the actual posture; When the fluctuation frequency is not within the preset frequency range, identifying whether there is a high-risk area on the cleaning path based on the environmental data collected by the perception module, and if there is no high-risk area on the cleaning path, controlling the photovoltaic robot to continue to execute the internal cleaning path; If there is a high-risk area on the cleaning path, the photovoltaic robot is controlled to bypass the high-risk area and continue to execute the internal cleaning path.

8. A photovoltaic robot, characterized in that: The photovoltaic robot comprises: A moving device for moving between photovoltaic panels; A detection device, installed on the mobile device, the detection device includes at least two perception modules, the perception modules are used to collect motion data or environmental data of the photovoltaic robot, and the perception modules include a visual perception module; A cleaning device, installed on the mobile device, for cleaning the photovoltaic panel; A control system, connecting the mobile device, the detection device and the cleaning device, is used to implement the path planning method of the photovoltaic robot as described in any one of claims 1 to 7.

9. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the path planning method for the photovoltaic robot as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor enables the processor to implement the path planning method for the photovoltaic robot according to any one of claims 1 to 7.

Images