Photovoltaic cleaning robot and positioning method and positioning system thereof

Abstract

The invention discloses a photovoltaic cleaning robot and a positioning method and a positioning system thereof. The positioning method comprises the following steps: S1, acquiring surrounding image data of the photovoltaic cleaning robot at a current position; s2, processing the surrounding image data, and identifying real-time texture information in an environment displayed by surrounding images; and S3, the real-time texture information is matched with preset texture information of the photovoltaic panel, and positioning information of the photovoltaic cleaning robot on the photovoltaic panel is obtained. According to the photovoltaic cleaning robot, the positioning technology of visual texture recognition is adopted, the structure is simple, assistance of an additional mechanical structure is not needed, the map construction cost is saved, the production and purchase cost is reduced, generalization of the cleaning robot is improved, and the photovoltaic cleaning robot can be applied to cleaning robots of different structures and different sizes and has good application prospects. And the device is also suitable for photovoltaic panels with different sizes and shapes.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic cleaning robot control, and in particular to a photovoltaic cleaning robot with a simple structure and easy use, and a positioning method and a positioning system thereof. Background Art

[0002] With the development of economy, solar energy is widely studied and used as a clean energy. Photovoltaic power generation is increasing year by year, and the photovoltaic power generation industry is getting more and more attention. In the research of photovoltaic power generation system, dust, bird droppings, sand and gravel attached to the surface of photovoltaic panels will affect the normal power generation of photovoltaic power generation system. Existing large-scale photovoltaic power stations generally install photovoltaic cleaning robots to automatically clean photovoltaic panels.

[0003] The existing positioning technologies of photovoltaic cleaning robots mainly include the following forms: (1) Mechanical positioning method: The robot is movably mounted on the surface of the photovoltaic panel by means of a mounting bracket or other structure, and the robot's movement and positioning are achieved through a mechanical structure. However, this method increases the structural complexity and inconvenience of the robot and is also costly. (2) Three-dimensional mapping method: This method uses a high-precision map and positioning device built in advance to achieve robot positioning, but the initial map building operation is cumbersome and costly; (3) Satellite positioning technology: This technology uses pre-planned areas and satellite signals to achieve robot positioning. However, it also has the problems of cumbersome initial operations and high costs. In addition, the accuracy and success rate of positioning depend on the strength of the satellite signal, and its reliability is slightly poor. Summary of the invention

[0004] The present invention provides a photovoltaic cleaning robot and a positioning method and a positioning system thereof, so as to solve the technical problems mentioned in the background technology.

[0005] In order to solve the above technical problems, the technical solution proposed by the present invention is: A positioning method for a photovoltaic cleaning robot comprises the following steps: S1, obtaining surrounding image data of the photovoltaic cleaning robot at the current position; S2, processing the surrounding image data to identify real-time texture information in the environment displayed by the surrounding image; S3. Match the real-time texture information with the preset texture information of the photovoltaic panel to obtain the positioning information of the photovoltaic cleaning robot on the photovoltaic panel.

[0006] As a further optimization of the above technical solution, in S3, the preset texture information of the photovoltaic panel is composed of a dark area and a light area, and the preset texture information is divided into: internal texture information, upper left corner texture information, lower left corner texture information, upper right corner texture information, lower right corner texture information, upper boundary texture information, lower boundary texture information, left boundary texture information and right boundary texture information according to the positional relationship and area proportion of the dark area and the light area. The preset texture information is the key to ensure the recognition success rate and positioning accuracy. When obtaining the preset texture information, the same equipment (equipment model, installation method, installation position, etc.) as the photovoltaic cleaning robot to obtain the surrounding image data of the current position is used to obtain the images of each corner, edge area and the internal area, and the same method is used for processing.

[0007] As a further optimization of the above technical solution, in S3, after obtaining the positioning information of the photovoltaic cleaning robot on the photovoltaic panel, a control instruction is generated according to the positioning information to adjust the motion state of the photovoltaic cleaning robot.

[0008] A control instruction is generated according to the positioning information. The specific method for adjusting the motion state of the photovoltaic cleaning robot is as follows: one of the corners of the photovoltaic panel is preset as the end angle area, the other corners and edges of the photovoltaic panel are preset as the non-end angle boundary, and the edge of the photovoltaic panel is preset as the end angle area. It is determined according to the positioning information whether the photovoltaic cleaning robot is in the preset end angle area. If the photovoltaic cleaning robot is in the end angle area, the cleaning is terminated. If the photovoltaic cleaning robot is not in the end angle area, it is determined whether the photovoltaic cleaning robot is in the non-end angle boundary. If so, a U-turn is performed and the straight cleaning is continued. If not, the straight cleaning is directly performed.

[0009] As a further optimization of the above technical solution, in S1, a camera unit installed on a photovoltaic cleaning robot is used to obtain the surrounding image data, and the camera unit is installed perpendicular to the plane of the photovoltaic panel to clearly obtain the edge, corner or internal area image of the photovoltaic panel.

[0010] As a further preferred embodiment of the above technical solution, in S2, the surrounding image data is subjected to grayscale conversion, noise reduction, binarization, and morphological filtering processing in sequence.

[0011] As a further optimization of the above technical solution, in S3, the real-time texture information is matched with the preset texture information of the photovoltaic panel according to the similarity, and the preset texture information most similar to the real-time texture information is selected, and the photovoltaic cleaning robot is located at the position corresponding to the preset texture information.

[0012] Based on the same technical concept, the present invention also provides a positioning system for a photovoltaic cleaning robot, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the above technical solution when executing the computer program.

[0013] Based on the same technical concept, the present invention also provides a photovoltaic cleaning robot, comprising a photovoltaic cleaning robot body, a camera unit, an execution unit and a positioning system of the photovoltaic cleaning robot described in the above technical solution; The camera unit is installed on the photovoltaic cleaning robot body, perpendicular to the plane of the photovoltaic panel, and is used to capture the surrounding image data of the photovoltaic cleaning robot at the current position and transmit it to the positioning system of the photovoltaic cleaning robot; The positioning system of the photovoltaic cleaning robot is used to process surrounding image data, obtain positioning information of the photovoltaic cleaning robot, generate control instructions, and transmit the control instructions to the execution unit; The execution unit is used to control the movement of the photovoltaic cleaning robot body and adjust the movement state of the photovoltaic cleaning robot according to the control instruction.

[0014] The present invention has the following beneficial effects: The photovoltaic cleaning robot of the present invention adopts the positioning technology of visual texture recognition, and has high positioning accuracy, so that the photovoltaic cleaning robot can adaptively clean on the panel. It has a simple structure, does not require the assistance of additional mechanical structure, and does not need to collect panel information in advance every time, which reduces the cost of use and improves the generalization of the cleaning robot. It can be applied to cleaning robots of different structures and sizes, and is also suitable for photovoltaic panels of different sizes and shapes, and has strong stability in use.

[0015] The present invention will be further described in detail below with reference to specific implementation modes. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a schematic diagram of a flow chart of a positioning method for a photovoltaic cleaning robot according to Example 1; Figure 2 This is a schematic diagram of preset texture information of a photovoltaic panel in the positioning method of a photovoltaic cleaning robot in Example 1; Figure 3 This is a schematic diagram of the graphic composition of the preset texture information in the positioning method of the photovoltaic cleaning robot in Example 1; Figure 4 This is a schematic diagram of the surrounding image data of the photovoltaic cleaning robot at the left border obtained in Example 1 and the processed image data and texture information; Figure 5This is a schematic diagram of the surrounding image data, processed image data and texture information of the photovoltaic cleaning robot in the upper right corner obtained in Example 1; Figure 6 This is a specific operation flow chart of S3 in the positioning method of the photovoltaic cleaning robot in Example 1; Figure 7 A path diagram of the photovoltaic cleaning robot in the photovoltaic cleaning robot positioning method of Example 1; Figure 8 This is a schematic diagram of the structure of the photovoltaic cleaning robot of Example 1. DETAILED DESCRIPTION

[0017] The embodiments of the present invention are described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.

[0018] Embodiment 1: The positioning method of the photovoltaic cleaning robot of this embodiment is as follows: Figure 1 As shown, the following steps are included: S1. Figure 2 The different matching areas shown in the figure obtain the preset texture information of the photovoltaic panel in advance, and determine the size of the black area according to the camera image that needs to be judged as arriving when the machine itself reaches the boundary, that is, the size of the binary area in the x-direction and y-direction, such as Figure 3 As shown, it includes internal texture information, upper left corner texture information, lower left corner texture information, upper right corner texture information, lower right corner texture information, upper boundary texture information, lower boundary texture information, left boundary texture information and right boundary texture information. Each preset texture information is composed of a dark area and a light area. The position relationship and area proportion of the dark area and the light area in the preset texture information of different areas are different. Figure 3 As shown; S2, obtain the surrounding image data of the photovoltaic cleaning robot at the current position, such as Figure 4 a (left border) and Figure 5 a (upper right corner); S3. Process the surrounding image data. Conventional image processing methods can be selected, including but not limited to grayscale, noise reduction, morphological filtering, grayscale-based threshold segmentation, and deep learning neural network segmentation methods to separate the foreground (photovoltaic panel) and background (non-photovoltaic panel area). The background is processed as a white area and the foreground area is retained, such as Figure 4 b and Figure 5 As shown in b, the real-time texture information in the environment displayed by the surrounding image is identified, such as Figure 4 c and Figure 5 c. Specific operation procedures such as Figure 6 As shown; S4, performing similarity matching between the real-time texture information and the preset texture information of the photovoltaic panel to obtain the positioning information of the photovoltaic cleaning robot on the photovoltaic panel; S5. Based on the positioning information, control instructions are formed to guide the photovoltaic cleaning robot to perform movement operations such as moving forward, turning, turning around, and ending. Figure 1 As shown, a control instruction is generated according to the positioning information, and a specific method for adjusting the motion state of the photovoltaic cleaning robot is as follows: one of the corners of the photovoltaic panel is preset as the end angle area, the other corners and edges of the photovoltaic panel are preset as the non-end angle boundary, and the edge of the photovoltaic panel is preset as the end angle area. It is determined according to the positioning information whether the photovoltaic cleaning robot is in the preset end angle area. If the photovoltaic cleaning robot is in the end angle area, the cleaning is terminated. If the photovoltaic cleaning robot is not in the end angle area, it is determined whether the photovoltaic cleaning robot is in the non-end angle boundary. If so, a U-turn operation is performed and the straight cleaning is continued. If not, the straight cleaning is directly performed. As in this embodiment, Figure 7 As shown, when the robot is placed at position 0 and starts to move forward, the matching texture is the lower boundary texture (such as position 1). When it reaches the lower right corner texture (position 2), it starts to turn around and the texture continues to move forward (position 3). When it finally reaches the upper left corner texture, the cleaning task is stopped.

[0019] The photovoltaic cleaning robot positioning system of this embodiment includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, steps S1 to S5 of the positioning method of this embodiment are implemented.

[0020] The photovoltaic cleaning robot of this embodiment is as follows: Figure 8 As shown, the photovoltaic cleaning robot body, the camera unit, the execution unit and the positioning system of the photovoltaic cleaning robot of this embodiment ( Figure 8 Wherein x is the photovoltaic cleaning robot body, and y is the camera unit); the camera unit is installed on the photovoltaic cleaning robot body, perpendicular to the plane of the photovoltaic panel, and is used to capture the surrounding image data of the photovoltaic cleaning robot at the current position and transmit it to the positioning system of the photovoltaic cleaning robot; the positioning system of the photovoltaic cleaning robot is used to process the surrounding image data using the positioning method of this embodiment, obtain the positioning information of the photovoltaic cleaning robot, generate control instructions, and transmit the control instructions to the execution unit; the execution unit is used to control the movement of the photovoltaic cleaning robot body, and adjust the movement state of the photovoltaic cleaning robot according to the control instructions.

[0021] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. For those skilled in the art, improvements and changes obtained without departing from the technical concept of the present invention should also be regarded as the protection scope of the present invention.

[0022] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A positioning method for a photovoltaic cleaning robot, characterized in that: The following steps are involved: S1, obtaining surrounding image data of the photovoltaic cleaning robot at the current position; S2, processing the surrounding image data to identify real-time texture information in the environment displayed by the surrounding image; S3. Match the real-time texture information with the preset texture information of the photovoltaic panel to obtain the positioning information of the photovoltaic cleaning robot on the photovoltaic panel.

2. The method for positioning a photovoltaic cleaning robot according to claim 1, characterized in that: In S3, the preset texture information of the photovoltaic panel is composed of a dark area and a light area. According to the positional relationship and area proportion of the dark area and the light area, the preset texture information is divided into: internal texture information, upper left corner texture information, lower left corner texture information, upper right corner texture information, lower right corner texture information, upper boundary texture information, lower boundary texture information, left boundary texture information and right boundary texture information.

3. The method for positioning a photovoltaic cleaning robot according to claim 1, characterized in that: In S3, after the positioning information of the photovoltaic cleaning robot on the photovoltaic panel is obtained, a control instruction is generated according to the positioning information to adjust the motion state of the photovoltaic cleaning robot.

4. The method for positioning a photovoltaic cleaning robot according to claim 3, characterized in that: A control instruction is generated according to the positioning information. The specific method for adjusting the motion state of the photovoltaic cleaning robot is as follows: one of the corners of the photovoltaic panel is preset as the end angle area, the other corners and edges of the photovoltaic panel are preset as the non-end angle boundary, and the edge of the photovoltaic panel is preset as the end angle area. It is determined according to the positioning information whether the photovoltaic cleaning robot is in the preset end angle area. If the photovoltaic cleaning robot is in the end angle area, the cleaning is terminated. If the photovoltaic cleaning robot is not in the end angle area, it is determined whether the photovoltaic cleaning robot is in the non-end angle boundary. If so, a U-turn is performed and the straight cleaning is continued. If not, the straight cleaning is directly performed.

5. The method for positioning a photovoltaic cleaning robot according to any one of claims 1 to 4, characterized in that: In S1, a camera unit installed on the photovoltaic cleaning robot is used to obtain the surrounding image data. The camera unit is installed perpendicular to the plane of the photovoltaic panel to clearly obtain the edge, corner or internal area image of the photovoltaic panel.

6. The photovoltaic cleaning robot positioning method according to any one of claims 1 to 4, characterized in that: In S2, the surrounding image data is processed in sequence by grayscale conversion, noise reduction, binarization, and morphological filtering.

7. The method for positioning a photovoltaic cleaning robot according to any one of claims 1 to 4, characterized in that: In S3, the real-time texture information is matched with the preset texture information of the photovoltaic panel according to the similarity, and the preset texture information most similar to the real-time texture information is selected, and the photovoltaic cleaning robot is located at the position corresponding to the preset texture information.

8. A positioning system for a photovoltaic cleaning robot, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

9. A photovoltaic cleaning robot, characterized in that: It comprises a photovoltaic cleaning robot body, a camera unit, an execution unit and the positioning system of the photovoltaic cleaning robot as described in claim 8; the camera unit is installed on the photovoltaic cleaning robot body, perpendicular to the plane of the photovoltaic panel, and is used to capture the surrounding image data of the photovoltaic cleaning robot at the current position and transmit it to the positioning system of the photovoltaic cleaning robot; the positioning system of the photovoltaic cleaning robot is used to process the surrounding image data, obtain the positioning information of the photovoltaic cleaning robot, generate control instructions, and transmit the control instructions to the execution unit; the execution unit is used to control the movement of the photovoltaic cleaning robot body, and adjust the movement state of the photovoltaic cleaning robot according to the control instructions.

Images