Posture detection method and posture detection system for photovoltaic panel

Abstract

The invention discloses an attitude detection method and system for a photovoltaic panel. The method comprises the following steps: S1, scanning through a laser radar to obtain light point cloud data; s2, segmenting the point cloud data through a clustering algorithm, and determining and storing clustering point cloud data to which the photovoltaic panel belongs; s3, performing plane fitting on the clustering point cloud data to which the photovoltaic panel belongs to obtain a plane equation, and projecting the point cloud data to the plane equation; s4, performing edge extraction on the point cloud of the photovoltaic panel to obtain a top edge linear equation; s5, postures are combined according to the vectors. According to the invention, the laser radar is used for data acquisition, the stability of equipment is improved, the operation is simple, other sensors are not depended on, a training set is not needed to train a model, the use stability and reliability are high, the system can be used in extreme environments such as night, heavy fog and rainy days, and the real-time analysis effect is achieved through algorithm optimization of point cloud processing. And the calculation efficiency of the attitude of the photovoltaic panel is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic devices, and particularly to a method and a system for detecting the attitude of a photovoltaic panel with simple operation and stable operation. Background Art

[0002] With the development of the economy, solar energy, as a clean energy source, has been widely studied and used. The power generation of photovoltaic power has been increasing year by year, and the photovoltaic power generation industry has attracted more and more attention. In the research of photovoltaic power generation systems, the installation angle of photovoltaic modules is one of the important factors for photovoltaic power generation. Therefore, how to accurately obtain the real-time angle of photovoltaic modules has a crucial impact on the normal use and fault troubleshooting of photovoltaic modules. The existing methods for detecting the attitude of photovoltaic panels mainly include two types. One is to install sensors to detect the tilt angle and direction, but this solution requires installing multiple sensors, with relatively high usage and maintenance costs. The other is to detect the attitude of photovoltaic panels through computer vision and image processing, but this solution requires a training set to train the model, with complex operations, and in weather conditions such as night, heavy rain, and heavy fog, image acquisition will be greatly affected, and the usage stability is not good. Summary of the Invention

[0003] The present invention provides a method and a system for detecting the attitude of a photovoltaic panel to solve the technical problems mentioned in the background art.

[0004] To solve the above technical problems, the technical solution proposed by the present invention is as follows: A method for detecting the attitude of a photovoltaic panel, comprising the following steps: S1. Obtain point cloud data of the area where the photovoltaic panel is located by lidar scanning; S2. Segment the point cloud data by a clustering algorithm, determine and save the clustered point cloud data belonging to the photovoltaic panel; S3. Perform plane fitting on the clustered point cloud data belonging to the photovoltaic panel to obtain a plane equation, project the clustered point cloud data belonging to the photovoltaic panel onto the plane equation, and use the projected data as the pure data of the photovoltaic panel; S4. Extract the edges of the pure data of the photovoltaic panel, extract the top edge for line fitting, and obtain the top edge line equation; S5. Calculate the plane normal vector according to the plane equation, calculate the top line segment vector according to the top line equation, intersect the plane normal vector and the top line segment vector to obtain the normal vector of the plane of these two vectors, and determine the attitude of the photovoltaic panel according to this normal vector.

[0005] As a further optimization of the above technical solution, in S1, the point cloud data is saved in the form of a data pool, and only the latest 3 frames of point cloud data are retained each time, and the final point cloud data is output after superposition. This operation can avoid the loss of calculation result accuracy caused by the loss of a certain frame of data, so as to obtain more information and more stable point cloud data.

[0006] As a further optimization of the above technical solution, in S1, according to the distance between the lidar and the photovoltaic panel, only the point cloud data around the photovoltaic panel is retained, and the point cloud data in this area is downsampled by a certain ratio to output the final point cloud data. This operation can facilitate the improvement of subsequent calculation efficiency.

[0007] As a further optimization of the above technical solution, in S2, after the point cloud data is segmented by a clustering algorithm, the center point coordinates of each cluster obtained by segmentation are calculated, and the clustered point cloud data of the photovoltaic panel is determined according to the relative clustering of the center point coordinates of each cluster and the lidar.

[0008] As a further optimization of the above technical solution, in S3, the calculation formula of the projection operation is P’ = P - (A * x + B * y + C * z + D) / (A ^2 + B ^2 + C ^2 ) * (A, B, C), where A, B, C, and D are the coefficients of the plane equation, and A, B, and C are the normal vectors; P is the point before the projection mapping, and P’ is the point after the projection mapping.

[0009] As a further optimization of the above technical solution, in S4, when extracting the edge of the pure data of the photovoltaic panel, the maximum Z value in the point cloud data is obtained by traversing, and all the point cloud data within 10 cm of the maximum Z value is selected as the top edge point cloud set for fitting the top edge straight line.

[0010] Based on the same inventive concept, the present invention also provides an attitude detection system for a photovoltaic panel, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method described in the above technical solution are implemented.

[0011] The present invention has the following beneficial effects: The present invention uses a lidar for data acquisition, achieving progress in equipment stability. It is simple to operate, does not rely on other sensors, and does not require a training set to train the model. It has high stability and reliability in use, can be used in extreme environments such as night, fog, and rain, and achieves real-time analysis through the optimization of the point cloud processing algorithm, improving the calculation efficiency of the attitude of the photovoltaic panel.

[0012] The following will further describe the present invention in detail with reference to specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is the point cloud data obtained by the lidar in Embodiment 1.

[0014] Figure 2 It is the pure data obtained after projecting the plane equation in Embodiment 1.

[0015] Figure 3 It is the display result of the photovoltaic panel attitude obtained by the attitude detection method in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following will describe the embodiments of the present invention in detail with reference to the drawings. However, the present invention can be implemented in many different ways defined and covered by the claims.

[0017] Embodiment 1: The attitude detection method of the photovoltaic panel in this embodiment includes the following steps: S1. Obtain the point cloud data of the area around the photovoltaic panel through the lidar. As Figure 1 shown, in the form of a data pool, 3 latest frames of point cloud data are retained each time in the array, and then multiple frames of point cloud data are superimposed and merged together and then point cloud analysis is performed to obtain more information and more stable point cloud data; S2. According to the approximate distance between the lidar and the photovoltaic panel, only retain the point cloud data within the target area and perform a certain proportion of downsampling on the data to facilitate the improvement of subsequent calculation efficiency; S3. Segment the point cloud data through a clustering algorithm (in this embodiment, the Euclidean clustering algorithm is used, and the segmentation criteria can be determined by setting relevant parameters such as the search radius size and the number of iterations of the Kdtree), calculate the center point coordinates of each cluster, and then judge the relative distance from the lidar to confirm the cluster to which the photovoltaic panel belongs and then only save this cluster; S4. Perform plane fitting on the photovoltaic panel cluster (in this embodiment, the RANSAC algorithm is used, and the fitting effect of the plane is adjusted by setting the maximum number of iterations and the maximum distance threshold). After obtaining the fitted plane equation, project the cluster point cloud data onto this plane equation. The calculation formula for the projection is the calculation formula for the projection P’ = P - (A * x + B * y + C * z + D) / (A ^2 + B ^2 + C ^2 ) * (A, B, C) to remove noise, and then save the data after projection as the pure data of the photovoltaic panel, as Figure 2 shown; S5. Perform edge extraction on the pure data of the photovoltaic panel, traverse to obtain the maximum Z value in the data, and then select all the point cloud data within 10 cm of this maximum value as the top edge point cloud set. Perform linear fitting on this top edge point cloud set to obtain the top straight line equation; S6. Calculate the plane normal vector according to the plane equation in S4, calculate the top line segment vector according to the top straight line equation in S5, and cross-multiply the plane normal vector and the top line segment vector to obtain the normal vector of the plane formed by these two vectors. As Figure 3 shown (in the figure, the red axis is the normal vector of the photovoltaic panel, the blue is the vector of the straight line equation, and the cross product of the two gives the attitude of the photovoltaic panel as Figure 3 shown), and this normal vector can represent the attitude of the photovoltaic panel.

[0018] The attitude detection system of the photovoltaic panel in this embodiment includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the attitude detection method of this embodiment is implemented.

[0019] The above is only the preferred embodiment 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 of this technology, the improvements and transformations obtained without departing from the technical concept of the present invention should also be regarded as the protection scope of the present invention.

[0020] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for detecting the posture of a photovoltaic panel, characterized in that: The following steps are involved: S1, obtaining point cloud data of the area where the photovoltaic panels are located through laser radar scanning; S2. Segmenting the point cloud data by a clustering algorithm, and determining and saving the clustered point cloud data to which the photovoltaic panel belongs; S3, performing plane fitting on the clustered point cloud data belonging to the photovoltaic panel to obtain a plane equation, projecting the clustered point cloud data belonging to the photovoltaic panel onto the plane equation, and using the projected data as pure data of the photovoltaic panel; S4, extracting the edge of the pure data of the photovoltaic panel, extracting the top edge for straight line fitting, and obtaining the straight line equation of the top edge; S5. Calculate the plane normal vector according to the plane equation, calculate the top line segment vector according to the top straight line equation, intersect the plane normal vector and the top line segment vector to obtain the normal vectors of the two vector planes, and determine the posture of the photovoltaic panel according to the normal vector.

2. The method for detecting the posture of a photovoltaic panel according to claim 1, characterized in that: In S1, the point cloud data is saved in a data pool, and only the latest 3 frames of point cloud data are retained each time, and the final point cloud data is output after superposition.

3. The method for detecting the posture of a photovoltaic panel according to claim 1, characterized in that: In S1, according to the distance between the lidar and the photovoltaic panel, only the point cloud data around the photovoltaic panel is retained, and the point cloud data in this area is downsampled by a certain proportion to output the final point cloud data.

4. The method for detecting the posture of a photovoltaic panel according to claim 1, characterized in that: In S2, after the point cloud data is segmented by a clustering algorithm, the center point coordinates of each cluster obtained by segmentation are calculated, and the cluster point cloud data of the photovoltaic panel is determined according to the center point coordinates of each cluster and the relative clustering of the laser radar.

5. The method for detecting the posture of a photovoltaic panel according to claim 1, characterized in that: In S3, the calculation formula of the projection operation is P'=P-(A*x+B*y+C*z+D) / (A ^2 +B ^2 +C ^2 )*(A, B, C), where A, B, C, D are the plane equation coefficients, A, B, C are the normal vectors; P is the point before projection mapping, and P' is the point after projection mapping.

6. The method for detecting the posture of a photovoltaic panel according to claim 1, characterized in that: In S4, when edge extraction is performed on the pure data of the photovoltaic panel, the maximum Z value in the point cloud data is traversed, and all point cloud data within the range of 10 cm of the maximum Z value are selected as the top edge point cloud set to fit the top edge straight line.

7. A photovoltaic panel posture detection system, 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 4 are implemented.

Images