Heliostat back surface type detection method and system

Abstract

The invention discloses a heliostat back surface type detection method and system, belongs to the technical field of heliostat surface type detection, and can solve the problems that an existing detection mode is low in detection efficiency, unstable in detection result and high in detection cost. The method comprises the following steps: S1, acquiring position information of a plurality of points on the back surface of a heliostat to be measured to form back surface point cloud data; the position information comprises an acquisition distance, a pitch angle and an azimuth angle; s2, performing coordinate transformation and rotation translation on the back point cloud data in sequence to obtain rotated point cloud data; the normal vector of the rotated point cloud data is parallel to the coordinate axis, and the central point of the rotated point cloud data coincides with the original point of the coordinate axis; and S3, fitting the back surface type of the heliostat to be measured according to the rotated point cloud data to obtain a back fitting surface type, and determining a back surface type detection result of the heliostat to be measured according to the back fitting surface type. The method is used for detecting the back surface type of the heliostat.

Description

Technical Field

[0001] The invention relates to a method and system for detecting the back surface shape of a heliostat, and belongs to the technical field of heliostat surface shape detection. Background Art

[0002] Solar thermal power generation uses heliostats to reflect sunlight and focus it on a tower to melt solid materials. The molten lava then heats water below the tower into steam to generate electricity. Solar thermal power generation requires many heliostats, and the curvature of each heliostat affects its ability to focus sunlight. Therefore, it is necessary to check the curvature of the heliostats.

[0003] Conventional technology typically uses manual dotting or full-row laser scanning to obtain point data on the backside of a heliostat. The backside profile is then fitted based on the acquired backside point data to verify the conformity of the heliostat's backside profile. However, manual dotting is inefficient and susceptible to human error, resulting in unstable test results. Full-row laser scanning, on the other hand, can result in asynchronous data processing and high costs. Summary of the Invention

[0004] The present invention provides a method and system for detecting the back surface shape of a heliostat, which can solve the problems of low detection efficiency, unstable detection results and high detection cost in existing detection methods.

[0005] In one aspect, the present invention provides a method for detecting the back surface shape of a heliostat, the method comprising:

[0006] S1, collecting the position information of multiple points on the back side of the heliostat to be measured to form back side point cloud data; the position information includes the collection distance, pitch angle and azimuth;

[0007] S2. performing coordinate transformation and rotational translation on the back point cloud data in sequence to obtain rotated point cloud data; wherein a normal vector of the rotated point cloud data is parallel to a coordinate axis, and a center point of the rotated point cloud data coincides with an origin of the coordinate axis;

[0008] S3. Fitting the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain a back surface fitting shape, and determining a back surface shape detection result of the heliostat to be measured according to the back surface fitting shape.

[0009] Optionally, the S2 specifically includes:

[0010] Performing coordinate transformation on the back point cloud data to obtain transformed point cloud data, and determining a normal vector of the transformed point cloud data using a principal component analysis method;

[0011] A rotation and translation matrix is ​​determined according to the normal vector, and the transformed point cloud data is rotated and translated according to the rotation and translation matrix to obtain rotated point cloud data.

[0012] Optionally, the determining the normal vector of the converted point cloud data by using a principal component analysis method specifically includes:

[0013] Performing feature centralization processing on the converted point cloud data to obtain centralized point cloud data;

[0014] A covariance matrix of the centered point cloud data is calculated, and a normal vector of the transformed point cloud data is determined according to the covariance matrix.

[0015] Optionally, determining the normal vector of the transformed point cloud data according to the covariance matrix is ​​specifically:

[0016] The eigenvalues ​​and eigenvectors of the covariance matrix are calculated, and the eigenvector corresponding to the minimum eigenvalue is used as the normal vector of the converted point cloud data.

[0017] Optionally, performing rotation and translation on the transformed point cloud data according to the rotation and translation matrix to obtain the rotated point cloud data is specifically:

[0018] The product of the rotation and translation matrix and the transformed point cloud data is calculated to obtain the rotated point cloud data.

[0019] Optionally, after S2, the method further includes:

[0020] interpolating the rotated point cloud data to obtain dense point cloud data;

[0021] Correspondingly, the back surface shape of the heliostat to be measured is fitted according to the rotated point cloud data in S3 to obtain the back surface fitting shape, which is specifically:

[0022] The back surface shape of the heliostat to be measured is fitted according to the dense point cloud data to obtain the back surface fitting shape.

[0023] Optionally, interpolation is performed on the rotated point cloud data to obtain dense point cloud data, specifically:

[0024] The rotated point cloud data is interpolated using a cubic spline linear interpolation method to obtain dense point cloud data.

[0025] Optionally, the S1 specifically includes:

[0026] Collecting position information of three reference points on the back of the heliostat to be measured; the three reference points are respectively close to three corners of the heliostat to be measured;

[0027] The scanning area of ​​the back side of the heliostat to be measured is determined according to the position information of the three reference points, and the points within the scanning area are scanned to obtain the point cloud data of the back side of the heliostat to be measured.

[0028] In another aspect, the present invention provides a heliostat backside profile detection system, the system comprising:

[0029] An acquisition unit is used to acquire position information of multiple points on the back side of the heliostat to be measured, and form back side point cloud data; the position information includes acquisition distance, pitch angle and azimuth angle;

[0030] a processing unit configured to sequentially perform coordinate conversion and rotational translation on the back point cloud data to obtain rotated point cloud data; wherein a normal vector of the rotated point cloud data is parallel to a coordinate axis, and a center point of the rotated point cloud data coincides with an origin of the coordinate axis;

[0031] The fitting unit is used to fit the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain a back fitting surface shape, and determine the back surface shape detection result of the heliostat to be measured according to the back fitting surface shape.

[0032] Optionally, the acquisition unit includes:

[0033] PTZ;

[0034] A laser rangefinder, mounted on the platform, is used to measure the distance between the laser rangefinder and multiple points on the back of the heliostat to be measured;

[0035] A driving motor connected to the pan / tilt platform for driving the pan / tilt platform to rotate horizontally or tilt;

[0036] The control unit is connected to the drive motor and the laser rangefinder, and is used to control the operation of the drive motor and obtain the pitch angle, horizontal rotation angle and measured distance of the laser rangefinder to form the back point cloud data.

[0037] The beneficial effects that the present invention can produce include:

[0038] The present invention provides a method and system for detecting the back surface shape of a heliostat. By collecting point cloud data of the back surface of the heliostat, the back surface point cloud data is sequentially subjected to coordinate transformation and rotational translation to obtain rotated point cloud data whose normal vector is parallel to the coordinate axis and whose center point coincides with the origin of the coordinate axis. The rotated point cloud data can more accurately simulate the back surface shape of the heliostat to be measured, thereby obtaining a more accurate back surface shape detection result, thereby improving detection efficiency.

[0039] The present invention provides a method and system for detecting the backside profile of a heliostat. Using a laser rangefinder, the system collects the distance, pitch angle, and azimuth angles of multiple points on the backside of the heliostat to be measured. This automated dot-marking technology improves the efficiency and accuracy of profile detection. Furthermore, it simplifies the detection mechanical structure and enhances the accuracy and efficiency of data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A flow chart of a method for detecting the back surface shape of a heliostat provided in an embodiment of the present invention;

[0041] Figure 2 A schematic diagram of the principle of principal component analysis in two-dimensional space provided by an embodiment of the present invention;

[0042] Figure 3 A schematic diagram of the principle of principal component analysis in three-dimensional space provided by an embodiment of the present invention;

[0043] Figure 4 This is a schematic diagram of the structure of the acquisition unit provided in an embodiment of the present invention.

[0044] Reference numerals:

[0045] 1. Pan / tilt head; 2. Laser rangefinder; 3. Pitch axis motor; 4. Horizontal axis motor; 5. Tripod bracket. DETAILED DESCRIPTION

[0046] The present invention is described in detail below with reference to the embodiments, but the present invention is not limited to these embodiments.

[0047] The embodiment of the present invention provides a method for detecting the back surface of a heliostat. Figure 1 As shown, the method includes:

[0048] S1. Collect the position information of multiple points on the back of the heliostat to be measured to form back point cloud data; the position information includes the collection distance, pitch angle and azimuth angle.

[0049] Specifically include:

[0050] (1) Collect the position information of three reference points on the back of the heliostat to be measured; the three reference points are respectively close to the three corners of the heliostat to be measured.

[0051] In practical applications, the three reference points may be located at the upper right, upper left, and lower left of the back of the heliostat, respectively.

[0052] In the embodiment of the present invention, the laser rangefinder 2 can be used to measure the position information of multiple points on the back of the heliostat to be measured.

[0053] (2) The scanning area of ​​the back side of the heliostat to be measured is determined based on the position information of the three reference points, and the points within the scanning area are scanned to obtain the point cloud data of the back side of the heliostat to be measured.

[0054] The laser rangefinder 2 is aligned with the first reference point, and then moves along a predetermined trajectory, automatically scanning from right to left and from top to bottom to obtain the point cloud data of the back side of the heliostat to be measured.

[0055] S2. Perform coordinate transformation and rotation and translation on the back point cloud data in sequence to obtain rotated point cloud data; the normal vector of the rotated point cloud data is parallel to the coordinate axis, and the center point of the rotated point cloud data coincides with the origin of the coordinate axis.

[0056] Specifically include:

[0057] (1) Perform coordinate transformation on the back point cloud data to obtain the transformed point cloud data, and use the principal component analysis method to determine the normal vector of the transformed point cloud data.

[0058] The collected back point cloud data is point data in 3D polar coordinates. Each point consists of three data: acquisition distance, pitch angle, and azimuth. Since this coordinate system is inconvenient for calculation, it is necessary to first perform coordinate conversion, that is, convert the original back point cloud data from polar coordinates to Cartesian coordinates. The conversion formula is as follows:

[0059]

[0060] Among them, (x, y, z) is the converted data, r is the acquisition distance, θ is the azimuth angle, and φ is the pitch angle.

[0061] After coordinate transformation, to facilitate calculation and analysis, the transformed point cloud data must be rotated and translated as a whole, keeping the normal vector of the point cloud plane parallel to the coordinate axes of the coordinate system and the center point of the transformed point cloud data at the origin of the coordinate system. Therefore, it is necessary to first calculate the normal vector of the point cloud, then calculate the rotation and translation matrix and multiply it with the transformed point cloud data to obtain the rotated point cloud data.

[0062] The PCA principal component analysis method can be used to calculate the normal vector of the transformed point cloud data. PCA is a mathematical transformation method that uses the idea of ​​dimensionality reduction to maintain the total variance of the variables during the transformation. It linearly transforms a given set of variables into another set of unrelated variables, maximizing the variance of the first variable after the transformation, i.e., the first principal component, and decreasing the variances of the other components.

[0063] Take two-dimensional space as an example, Figure 2As shown in the figure, the first straight line (y' axis) represents the first principal component, which is at an angle to the original x-axis. This angle indicates the direction of the first principal component, which captures the maximum variance of the data. The second straight line (x' axis) is perpendicular to the first line and represents the second principal component, which captures the remaining variance. The dots in the figure are unevenly distributed in the original coordinate system (x, y), but appear more evenly distributed in the new coordinate system (x', y'). This shows that PCA successfully transforms the data into a new space, making the data more evenly distributed on the new coordinate axis, making it easier to perform subsequent analysis or visualization.

[0064] In the present invention, the converted point cloud data is a set of three-dimensional point coordinates, whose variables are the three coordinate values ​​of X, Y, and Z. After the transformation, there should be three principal components. For a spatial plane, the point set distribution is most discrete in the direction parallel to the plane, with the largest variance. In the direction perpendicular to the plane, the point set distribution is most concentrated, with the smallest variance. That is, the third principal component of the spatial plane is the vector perpendicular to the spatial plane. Since the most critical part of plane fitting is the fitting of the normal vector, PCA is used to obtain the third principal component of the point set, which can further fit the plane equation, such as Figure 3 shown.

[0065] For the transformed point cloud data to be fitted in the coordinate system XYZ, the principal component analysis method is used to analyze it, and three eigenvalues ​​λ1, λ2, and λ3 arranged from large to small can be obtained. The corresponding principal components are V1, V2, and V3, respectively. V1 and V2 form a set of bases of the plane to be fitted, and V3 is orthogonal to V1 and V2 and is the normal vector perpendicular to the plane to be fitted. Figure 3 As shown in the figure, after principal component analysis, the three principal components V1, V2, and V3 of the point cloud in the XYZ coordinate system form the three bases of the new coordinate system X'Y'Z'. V1 and V2 are a set of bases of the plane X'O'Z', and V3 is the basis in the O'Z' direction, that is, the normal vector of the point cloud data after transformation.

[0066] Therefore, the above-mentioned principal component analysis method is used to determine the normal vector of the transformed point cloud data, specifically including:

[0067] (a) Perform feature centering processing on the converted point cloud data to obtain centralized point cloud data.

[0068] The mean of each dimension is subtracted from the data of that dimension, and after the transformation, the mean of each dimension becomes zero.

[0069] The point set P after feature centering is shown in formula (2).

[0070]

[0071] Among them, x i 、yi 、z i are the coordinates of the centered point.

[0072] (b) Calculate the covariance matrix of the centered point cloud data.

[0073] Calculate the covariance matrix of the three coordinates. The covariance matrix C is as follows:

[0074]

[0075] Where cov(x,y) is the covariance of the x-coordinate and the y-coordinate, and cov(x,x) is the variance of the x-coordinate. The covariance calculation formula is shown in formula (4).

[0076]

[0077] Among them, x i 、y i are the coordinates of the centered point.

[0078] When the covariance is greater than zero, it means that x and y are positively correlated. When the covariance is less than zero, x and y are negatively correlated. When the covariance is zero, x and y are independent of each other.

[0079] (c) Calculate the eigenvalues ​​and eigenvectors of the covariance matrix, and use the eigenvector corresponding to the smallest eigenvalue as the normal vector of the transformed point cloud data.

[0080] Calculate the eigenvalues ​​and eigenvectors of the covariance matrix C. The calculated eigenvalues ​​are sorted from largest to smallest as λ1, λ2, and λ3, and their corresponding eigenvectors are ξ1, ξ2, and ξ3. Obviously, the eigenvectors ξ1 and ξ2 corresponding to the two larger λs are a set of bases for the point cloud surface, and ξ3 is the normal vector of the point cloud surface, that is, the normal vector of the transformed point cloud data.

[0081] When using PCA to calculate the normal vector of a nearly planar surface, the results can be calculated well even in the presence of noise points. This is because in a planar point cloud, the distance that the noise points deviate from the plane is small relative to the extent of the plane, and their impact on the results can be ignored.

[0082] (2) Determine the rotation and translation matrix based on the normal vector, and rotate and translate the transformed point cloud data according to the rotation and translation matrix to obtain the rotated point cloud data.

[0083] After obtaining the normal vector of the transformed point cloud data, the rotation and translation matrix is ​​determined according to the normal vector, and then the product of the rotation and translation matrix and the transformed point cloud data is calculated to obtain the rotated point cloud data.

[0084] S3. Fitting the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain a back surface fitting shape, and determining a back surface shape detection result of the heliostat to be measured according to the back surface fitting shape.

[0085] Furthermore, after S2, the method further includes:

[0086] The rotated point cloud data is interpolated to obtain dense point cloud data.

[0087] Specifically: the rotated point cloud data is interpolated using the cubic spline linear interpolation method to obtain dense point cloud data.

[0088] The rotated point cloud data is linearly interpolated using cubic spline interpolation. Cubic spline interpolation is a commonly used numerical analysis method that can fit a smooth, continuous function curve to a given set of scattered data points. The basic idea is to approximate a small interval of data using low-degree polynomials and utilize the connection conditions at the junctions of these polynomials to ensure the smoothness of the entire curve. It is widely used in data approximation and interpolation problems in engineering, science, and mathematics.

[0089] Definition: For n known data points (x i ,y i ), where x1 <x2<...<x n , the cubic spline interpolation method can find a set of functions S(x) that meet the following conditions: (1) S(x) is a cubic polynomial in each small interval; (2) S(x) has the same first-order and second-order derivatives at the connection of two adjacent small intervals, that is, S'(x i )=S'(x i +1), S”(x i )=S”(x i +1); (3) S(x) passes through all given data points, that is, S(x i )=y i .

[0090] The main steps of the cubic spline linear interpolation method are as follows: first, fit a cubic polynomial in each small interval; then set the boundary conditions to obtain a system of equations; then solve the system of equations to find the coefficients of each small interval; finally, give the interpolation points and use the coefficients to find the corresponding function values.

[0091] Correspondingly, in S3, the back surface shape of the heliostat to be measured is fitted according to the rotated point cloud data to obtain the back surface fitting shape, which is specifically:

[0092] The back surface shape of the heliostat to be measured is fitted according to the dense point cloud data to obtain the back surface fitting shape.

[0093] After using cubic spline linear interpolation to generate dense point cloud data, the back surface profile of the heliostat to be measured is fitted using this dense point cloud data to obtain a fitted back surface profile. This back surface profile is then compared with the standard back surface profile of the heliostat to obtain a back surface profile test result. If the test result fails, the back surface profile of the heliostat to be measured is adjusted, and the position information of the back points is collected again. Coordinate transformation and rotation and translation are performed again, and the back surface profile fitting is then repeated until the test result is qualified.

[0094] The present invention realizes the transformation of data points from three-dimensional polar coordinates to Cartesian coordinates, and then realizes heliostat mirror surface profile fitting through PCA principal component analysis method, providing a basis for surface profile detection.

[0095] Another embodiment of the present invention provides a heliostat back surface shape detection system, the system comprising:

[0096] The acquisition unit is used to collect the position information of multiple points on the back of the heliostat to be measured to form back point cloud data; the position information includes the acquisition distance, pitch angle and azimuth angle;

[0097] The processing unit is used to perform coordinate conversion and rotation and translation on the back point cloud data in sequence to obtain rotated point cloud data; the normal vector of the rotated point cloud data is parallel to the coordinate axis, and the center point of the rotated point cloud data coincides with the origin of the coordinate axis;

[0098] The fitting unit is used to fit the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain the back fitting surface shape, and determine the back surface shape detection result of the heliostat to be measured according to the back fitting surface shape.

[0099] The acquisition unit includes:

[0100] PTZ 1;

[0101] The laser rangefinder 2 is mounted on the gimbal 1 and is used to measure the distance between it and multiple points on the back of the heliostat to be measured;

[0102] A drive motor connected to the pan / tilt head 1 for driving the pan / tilt head 1 to rotate horizontally or tilt;

[0103] The control unit is connected to the drive motor and the laser rangefinder 2, and is used to control the operation of the drive motor and obtain the pitch angle, horizontal rotation angle and measured distance of the laser rangefinder 2 to form the back point cloud data.

[0104] The control unit includes a host computer and a single-chip microcomputer, and the data transmission can be realized by using a Bluetooth module.

[0105] refer to Figure 4As shown, the laser rangefinder 2 measures the distance between the laser and the mirror and sends the laser data back to the microcontroller through the CAN bus.

[0106] The pan-tilt platform 1 is used to carry the laser rangefinder 2 to perform scanning activities in the horizontal and vertical directions.

[0107] The driving motor includes a pitch axis motor 3 and a horizontal axis motor 4. The pitch axis motor 3 drives the pan-tilt platform 1 to complete a vertical scanning action; the horizontal axis motor 4 drives the pan-tilt platform 1 to complete a horizontal scanning action.

[0108] The single chip microcomputer controls the scanning action of the pitch motor and the horizontal motor, reads the laser rangefinder 2 and the rotation angles in the pitch and horizontal directions, controls the display screen to display the operation interface, and controls the Bluetooth module to upload data to the host computer.

[0109] The host computer receives data from the microcontroller.

[0110] Furthermore, a tripod bracket 5 may be provided below the pan-tilt platform 1 so as to conveniently adjust the collection height of the laser rangefinder 2 .

[0111] The specific operating process is as follows: First, the system is powered on, and the microcontroller initializes the drive motor, laser rangefinder 2, display, and Bluetooth module. The user then controls the pan / tilt head 1 using buttons on the LCD screen. Each button corresponds to a callback function, which calls the CAN bus command sending function to control the movement of the drive motor. Three key reference points are collected sequentially: these are located at the upper right, upper left, and lower left of the heliostat. The data for each reference point includes the distance from the point to the laser rangefinder 2, the pitch angle, and the horizontal angle. This data is stored in a structure array for easy access later. Next, the auto-scan button is clicked. The laser rangefinder 2 is aligned with the first reference point, and the pan / tilt head 1 then automatically scans from right to left and top to bottom, following a predetermined trajectory. Finally, the microcontroller controls the Bluetooth module to transmit the collected data to the host computer via the serial port protocol.

[0112] The acquisition unit of the present invention realizes the movement of the laser rangefinder 2 along a predetermined trajectory through the laser rangefinder 2, the pan-tilt platform 1, the drive motor, and the control unit, thereby completing automatic scanning of the heliostat back surface profile data and data transmission processing, thereby improving the heliostat surface profile detection efficiency.

[0113] The present invention provides a method and system for detecting the back surface shape of a heliostat. By collecting point cloud data of the back surface of the heliostat, the back surface point cloud data is sequentially subjected to coordinate transformation and rotational translation to obtain rotated point cloud data whose normal vector is parallel to the coordinate axis and whose center point coincides with the origin of the coordinate axis. The rotated point cloud data can more accurately simulate the back surface shape of the heliostat to be measured, thereby obtaining a more accurate back surface shape detection result, thereby improving detection efficiency.

[0114] The present invention provides a method and system for detecting the backside profile of a heliostat. Using a laser rangefinder 2, the system collects the distance, elevation, and azimuth angles of multiple points on the backside of the heliostat to be measured. This automated dot-marking technology improves the efficiency and accuracy of profile detection. Furthermore, it simplifies the detection mechanical structure and enhances the accuracy and efficiency of data processing.

[0115] The above descriptions are merely a few embodiments of the present application and do not constitute any form of limitation to the present application. Although the present application discloses the preferred embodiments as above, they are not intended to limit the present application. Any technical personnel familiar with the present profession, without departing from the scope of the technical solution of the present application, using the technical content disclosed above to make slight changes or modifications are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for detecting the back surface of a heliostat, characterized in that: The method comprises: S1, collecting the position information of multiple points on the back side of the heliostat to be measured to form back side point cloud data; the position information includes the collection distance, pitch angle and azimuth; S2, performing coordinate transformation and rotational translation on the back point cloud data in sequence to obtain rotated point cloud data; the normal vector of the rotated point cloud data is parallel to the coordinate axis, and the center point of the rotated point cloud data coincides with the origin of the coordinate axis; S3. Fitting the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain a back surface fitting shape, and determining a back surface shape detection result of the heliostat to be measured according to the back surface fitting shape.

2. The method according to claim 1, characterized in that The S2 specifically includes: Performing coordinate transformation on the back point cloud data to obtain transformed point cloud data, and determining the normal vector of the transformed point cloud data by using a principal component analysis method; A rotation and translation matrix is ​​determined according to the normal vector, and the transformed point cloud data is rotated and translated according to the rotation and translation matrix to obtain rotated point cloud data.

3. The method according to claim 2, characterized in that The method of determining the normal vector of the converted point cloud data by using the principal component analysis method specifically includes: Performing feature centralization processing on the converted point cloud data to obtain centralized point cloud data; A covariance matrix of the centered point cloud data is calculated, and a normal vector of the transformed point cloud data is determined according to the covariance matrix.

4. The method according to claim 3, characterized in that: The determining of the normal vector of the transformed point cloud data according to the covariance matrix is ​​specifically: The eigenvalues ​​and eigenvectors of the covariance matrix are calculated, and the eigenvector corresponding to the smallest eigenvalue is used as the normal vector of the converted point cloud data.

5. The method according to claim 2, characterized in that: The step of performing rotation and translation on the transformed point cloud data according to the rotation and translation matrix to obtain the rotated point cloud data is specifically as follows: The product of the rotation and translation matrix and the transformed point cloud data is calculated to obtain the rotated point cloud data.

6. The method according to claim 1, characterized in that After S2, the method further includes: Interpolating the rotated point cloud data to obtain dense point cloud data; Correspondingly, the back surface shape of the heliostat to be measured is fitted according to the rotated point cloud data in S3 to obtain the back surface fitting shape, which is specifically: The back surface shape of the heliostat to be measured is fitted according to the dense point cloud data to obtain the back surface fitting shape.

7. The method according to claim 6, characterized in that The rotated point cloud data is interpolated to obtain dense point cloud data, specifically: The rotated point cloud data is interpolated using a cubic spline linear interpolation method to obtain dense point cloud data.

8. The method according to claim 1, characterized in that: The S1 specifically includes: Collecting position information of three reference points on the back of the heliostat to be measured; the three reference points are respectively close to three corners of the heliostat to be measured; The scanning area of ​​the back side of the heliostat to be measured is determined according to the position information of the three reference points, and the points within the scanning area are scanned to obtain the back side point cloud data of the heliostat to be measured.

9. A heliostat back surface detection system, characterized in that: The system comprises: A collection unit is used to collect the position information of multiple points on the back side of the heliostat to be measured to form back side point cloud data; the position information includes the collection distance, pitch angle and azimuth angle; A processing unit, used to perform coordinate transformation and rotational translation on the back point cloud data in sequence to obtain rotated point cloud data; the normal vector of the rotated point cloud data is parallel to the coordinate axis, and the center point of the rotated point cloud data coincides with the origin of the coordinate axis; A fitting unit is used to fit the back surface shape of the heliostat to be measured according to the rotated point cloud data to obtain a back fitting surface shape, and determine a back surface shape detection result of the heliostat to be measured according to the back fitting surface shape.

10. The system according to claim 9, characterized in that The acquisition unit comprises: PTZ; A laser rangefinder, mounted on the pan-tilt platform, is used to measure the distance between the laser rangefinder and multiple points on the back of the heliostat to be measured; A driving motor connected to the pan / tilt platform, and used to drive the pan / tilt platform to rotate horizontally or tiltly; The control unit is connected to the driving motor and the laser rangefinder, and is used to control the driving motor to work, and obtain the pitch angle, horizontal rotation angle and measured distance of the laser rangefinder to form back point cloud data.

Images