A method for measuring the pointing error of a heliostat using moonlight and a camera

Abstract

The application belongs to the technical field of heliostat measurement, and particularly relates to a method for measuring pointing error of heliostat by using moonlight and a camera, comprising the following steps: rotating the heliostat, guiding the moonlight reflected by the heliostat to the camera; detecting the actual rotation angle of the heliostat when the center of the reflected light beam is located at the position of the camera; at the same time, calculating the theoretical rotation angle of the heliostat at the position based on the heliostat coordinates, the camera coordinates and the moonlight azimuth vector; the difference between the actual rotation angle and the theoretical rotation angle is the pointing error of the heliostat at the current position. The method has the advantages that the moon azimuth changes naturally with time, theoretically, infinite measurement azimuths can be provided, and an unmanned aerial vehicle or an additional calibration tower is not needed, so that the system cost is significantly reduced.

Description

Technical Field

[0001] This application belongs to the field of heliostat measurement technology, specifically relating to a method for measuring the pointing error of a heliostat using moonlight and a camera. Background Technology

[0002] In tower-type solar thermal systems, to ensure that heliostats have sufficient pointing accuracy so that the reflected sunlight can be continuously and stably focused on the receiver, a calibration system is typically required to continuously correct the pointing error of the heliostats. This calibration process usually includes three steps: pointing error measurement, error analysis, and daily angle correction. Pointing error measurement is the foundation of the calibration system and a key factor affecting calibration efficiency. Currently common pointing error measurement techniques mainly include: white target-based measurement techniques, solar light source and camera-based measurement techniques, and artificial light source and camera-based measurement techniques.

[0003] The white target-based measurement technique involves controlling a heliostat to guide the reflected light beam to a white target below the receiver. The heliostat is then rotated to align the center of the light spot with the target's center. The difference between the actual rotation angle of the heliostat and its theoretical rotation angle represents the pointing error at the current position. This technique offers high measurement accuracy, but its calibration efficiency is low because a single white target can only calibrate one heliostat at a time.

[0004] Measurement technology based on solar light source and camera: The heliostat projects a reflected light beam onto the camera, and machine vision technology is used to identify the actual rotation angle of the heliostat when the beam center is at the camera position. The deviation from the theoretical rotation angle is the pointing error. This technology can calibrate multiple heliostats simultaneously using a single camera, which is more efficient than the white target technology, and is therefore widely used in commercial projects. However, the superposition of multiple beams can easily lead to increased camera temperature. Even with dedicated protective covers and filters, the number of heliostats that a single camera can calibrate simultaneously is still limited, restricting the overall calibration efficiency.

[0005] Measurement technology based on artificial light sources and cameras: Its principle is similar to that of solar-powered camera technology, the difference being the use of artificial light sources instead of sunlight. Artificial light sources avoid camera overheating issues, thus achieving higher calibration efficiency. However, to provide sufficient light source location, the artificial light source typically needs to be mounted on a drone or deployed on multiple calibration towers, significantly increasing system costs. Summary of the Invention

[0006] This invention proposes a method for measuring the pointing error of a heliostat using moonlight and a camera. Moonlight, similar to artificial light sources, does not cause significant camera overheating, thus greatly improving efficiency compared to solar-based technologies. Furthermore, the moon's position changes naturally over time, theoretically providing an unlimited number of measurement positions, eliminating the need for drones or additional calibration towers, thereby significantly reducing system costs. The technical solution is as follows:

[0007] A method for measuring the pointing error of a heliostat using moonlight and a camera includes a heliostat and a camera. The heliostat is rotated to guide the reflected moonlight to the camera. The actual rotation angle of the heliostat is detected when the center of the reflected beam is located at the camera position. Simultaneously, the theoretical rotation angle of the heliostat reflecting moonlight to the camera is calculated based on a heliostat calibration model. The difference between the actual rotation angle and the theoretical rotation angle is the pointing error of the heliostat at the current position.

[0008] Preferably, the calculation of the moonlight azimuth vector is based on the geometric projection method:

[0009] A rectangular coordinate system is established with the observation point on Earth as the origin, i.e., the position of the heliostat is the origin, the X-axis points to geographic east, the Y-axis points to geographic north, and the Z-axis is perpendicular to the horizontal plane and pointing upwards. Based on the lunar and solar coordinates, a three-dimensional unit vector is calculated pointing from the origin to the center point of the area of ​​the moon illuminated by the sun; this vector is called the lunar azimuth vector and is denoted as . Simultaneously, the proportion of the moon's illuminated area to its maximum visible area is calculated; this is called the moonlight visible area proportion, denoted as . .

[0010] Preferably, by detecting the contour of the reflected beam, the actual rotation angle of the heliostat when the center of the reflected beam is located at the camera is calculated. The calculation steps for the actual rotation angle are as follows:

[0011] S1. Perform beam positioning;

[0012] S2. Identify the light spot, rotate the heliostat to continuously change the position of the reflected beam, so that the light spot image inside the camera keeps changing, record the maximum pixel value of the light spot image, and use the DBSCAN algorithm to aggregate the maximum pixel values ​​at different positions into light spot class and background class. The average of the minimum value of the light spot class and the maximum value of the background class is used as the brightness threshold to distinguish between the background and the light spot.

[0013] S3. Beam profile recognition: Rotate the heliostat to move the reflected beam in N directions around the camera's azimuth vector, and identify the boundaries of the beam in each direction.

[0014] S4. Beam center calculation: Calculate the convex hull of the beam profile boundary and use the center of the convex hull as the beam center.

[0015] S5. Calculation of actual rotation angle: Based on the corrected parameters of the heliostat model at the beam boundary, the rotation angle when the heliostat reflects moonlight to the center of the beam is calculated and taken as the actual rotation angle.

[0016] Preferably, the theoretical rotation angle of the heliostat is calculated using the current heliostat calibration model parameters, based on the moonlight azimuth vector and the camera azimuth vector. The difference between the theoretical rotation angle and the actual rotation angle is the heliostat pointing error.

[0017] Preferably, to determine whether the current moment allows for the measurement of the heliostat's pointing error using moonlight, the permitted time for measurement must simultaneously meet the following three conditions:

[0018] Solar altitude angle It must be less than the threshold. ;

[0019] Moon altitude angle It must be greater than the threshold. ;

[0020] Percentage of the area visible to the moonlight It must be greater than the threshold. .

[0021] Preferably, it is necessary to determine whether moonlight is visible before taking the measurement. The specific method is as follows:

[0022] choose n Each heliostat has been calibrated. The camera controls the heliostat to reflect moonlight onto the camera and identifies the reflected light spot of each heliostat. When the maximum brightness of the reflected light spots of all heliostats is greater than a given threshold, it means that the moonlight is visible.

[0023] Preferably, the beam positioning steps are as follows: calculate the imaging area of ​​the heliostat on the camera image, given the moonlight azimuth vector and the camera azimuth vector, calculate the theoretical rotation angle of the heliostat according to the correction model. Control the heliostat to rotate to If a moonlight spot can be identified within the imaging area, the beam positioning ends.

[0024] Preferably, if no light spot can be detected within the imaging area, then a set step size is used. The heliostat's rotation angle is adjusted according to the spiral trajectory. At each rotation position, a spot in the imaging area is detected. If a spot appears but its maximum brightness is less than a given threshold, the surrounding area is searched using this spot as the center. Then, the search is repeated in the direction with the highest brightness value in the surrounding area until the maximum brightness value of the spot is greater than the given threshold.

[0025] Compared with the prior art, the beneficial effects of this application are as follows:

[0026] (1) Compared with the traditional method of measuring the pointing error of a heliostat using solar light, the method of the present invention uses moonlight instead of solar light. Moonlight will not cause the camera to heat up, so the measurement efficiency is higher and there is no safety hazard.

[0027] (2) Compared with the traditional method of measuring the pointing error of a heliostat using artificial light sources, the method of the present invention uses moonlight instead of artificial light sources. Moonlight is a natural light source and its orientation changes constantly. It does not require additional calibration towers or light source carrier equipment such as drones, so the measurement cost is lower. Attached Figure Description

[0028] Figure 1 A schematic diagram for calculating the moonlight azimuth vector;

[0029] Figure 2 Workflow for measuring the pointing error of a heliostat using moonlight and a camera;

[0030] Figure 3 A schematic diagram illustrating the simultaneous identification of multiple heliostat moonlight spots;

[0031] Figure 4 For the beam positioning process;

[0032] Figure 5 This is a schematic diagram for beam contour recognition;

[0033] Figure 6 This is a rendering. Detailed Implementation

[0034] The technical solution of this application will be described in detail below with reference to specific embodiments and accompanying drawings. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof. Specific technical features can be combined with each other.

[0035] This invention proposes a method for measuring the pointing error of a heliostat using moonlight and a camera. The method includes: rotating the heliostat to guide the reflected moonlight to the camera; detecting the actual rotation angle of the heliostat when the center of the reflected beam is located at the camera position; simultaneously, calculating the theoretical rotation angle of the heliostat at that position based on the heliostat coordinates, camera coordinates, and the moonlight azimuth vector; the difference between the actual rotation angle and the theoretical rotation angle is the pointing error of the heliostat at the current position. This method is applicable to the calibration of heliostats and the evaluation of their tracking errors.

[0036] As attached Figure 1 As shown, the calculation of the moonlight azimuth vector is based on the geometric projection method. This method establishes a rectangular coordinate system with the observation point on Earth as the origin; that is, the heliostat position is the origin, the X-axis points to geographic east, the Y-axis points to geographic north, and the Z-axis is perpendicular to the horizontal plane and pointing upwards. Based on the lunar and solar coordinates, a three-dimensional unit vector pointing from the origin to the center point of the sun-illuminated area of ​​the moon is calculated. This vector is called the moonlight azimuth vector and is denoted as . Simultaneously, the proportion of the moon's illuminated area to its maximum visible area is calculated; this is called the moonlight visible area proportion, denoted as . .

[0037] The rotation angle of a heliostat is defined as the rotation angle of the heliostat's current position relative to a reference position. Taking a common horizontal and vertical dual-axis driven heliostat as an example, the rotation angle is a two-dimensional vector. It includes two rotational components: horizontal and pitch.

[0038] The camera azimuth vector is defined as the unit vector pointing from the heliostat position to the camera, denoted as . .

[0039] The heliostat calibration model defines the relationship between the moonlight azimuth vector, the heliostat rotation angle, and the camera azimuth vector. The specific expression of the calibration model is related to the heliostat's mechanical structure, error compensation algorithm, etc. This invention does not impose any special restrictions on the calibration model. The general calibration model is shown in Equations 1 and 2.

[0040] The positive correction model function of the heliostat can calculate the rotation angle of the heliostat when reflecting moonlight to the camera position based on the moonlight azimuth vector and the camera azimuth vector. It is the parameter vector used in the heliostat correction model, which usually includes heliostat coordinates, heliostat azimuth, error compensation parameters, etc. This represents the heliostat reverse correction model, which can calculate the camera azimuth vector based on the moonlight azimuth vector and the heliostat rotation angle.

[0041] （1）；

[0042] （2）；

[0043] The flowchart of the present invention is attached. Figure 2 As shown, the main process includes determining the measurement time, determining the moonlight visibility status, locating the beam, identifying the beam outline, and calculating the pointing error.

[0044] (1) Measurement time judgment: Determine whether the current time allows the use of moonlight to measure the pointing error of the heliostat. The time allowed for measurement work must simultaneously meet the following 3 conditions, see Formula 3: For a given time Solar altitude angle It must be less than the threshold. lunar altitude angle It must be greater than the threshold. Percentage of the visible moonlight area It must be greater than the threshold. .when The time t represents a given moment that allows the heliostat pointing error to be measured using moonlight.

[0045] (3).

[0046] (2) Determining Moonlight Visibility: Cloudy weather can cause moonlight to dim or even become invisible, leading to measurement failure or incorrect results. Therefore, it is necessary to determine whether moonlight is visible before measurement. The specific method is to select... n Several calibrated heliostats were used to control the reflection of moonlight onto the camera throughout the testing process. The reflected light spots from each heliostat were identified at regular intervals, allowing for precise control of the light at any given moment. Let the maximum brightness of a single light spot be denoted as . When the maximum brightness of all heliostat spots is greater than If the value is 0, it means that moonlight is currently visible. The calculation method is shown in Formula 4.

[0047] (4).

[0048] (4) Beam positioning: The beam positioning process involves moving the reflected moonlight beam to the camera position so that the moonlight spot can be identified on the camera screen. The specific process is as follows:

[0049] 1) Based on the heliostat coordinates and the camera imaging model, calculate the position of the heliostat in the camera image. Divide a rectangular area centered on this position as the imaging area of ​​the heliostat within the camera. Heliostats with non-overlapping imaging areas can independently and simultaneously detect pointing errors.

[0050] 2) According to Formula 1, given the camera orientation vector The azimuth vector of moonlight at a given moment The theoretical rotation angle of the heliostat was calculated. .

[0051] 3) Control the heliostat to rotate to If the imaging area can detect moonlight spots, and the maximum brightness of the spots is greater than a given threshold... Then the beam positioning is complete, see appendix. Figure 3 As shown.

[0052] 4) If no light spot can be detected within the imaging area, then proceed with a certain step size. Modify the heliostat rotation angle according to the spiral trajectory, where Indicates the step size for modification in the horizontal direction. This indicates the step size for adjusting the pitch direction. See appendix. Figure 4 As shown, a light spot is detected within the imaging area after each rotation. If a light spot appears, but its maximum brightness is less than a given threshold, the system will detect it. Then, using this as the center, search the surrounding area, and move towards the direction with the highest brightness value, repeating the search process until the maximum brightness value of the spot is greater than a given threshold. Let the rotation angle of the effective light spot found be denoted as . The corresponding camera orientation vector is calculated using Formula 2. .

[0053] (5) Moonlight Spot Imaging Analysis: Because moonlight is less bright than sunlight, the brightness of moonlight spots is easily affected by specular reflectivity and the distance between the heliostat and the camera. To improve measurement accuracy, it is necessary to analyze the moonlight spots to determine more suitable spot recognition parameters. The specific method is as follows: Rotate the heliostat to continuously change the angle of the reflected beam, so that the spot imaging in the camera continuously changes, and record the maximum pixel value of the imaging area. As shown in Formula 5, the DBSCAN algorithm is used to aggregate the maximum pixel values ​​of the imaging corresponding to the reflected beams at different angles into spot classes. and background class If the profile coefficient This indicates that the grouping of the light spot class and the background class is reliable, with the background class being the most reliable group. Maximum value and spot type The mean of the minimum value Replace the initial spot brightness detection threshold This serves as a brightness threshold for subsequently distinguishing between the background and the light spot.

[0054] (5)

[0055] (6).

[0056] (6) Beam profile recognition: Step 4 initially determined the camera's orientation vector. Rotate the heliostat to make the reflected beam of light Moving in N directions from the center, identify the boundaries of the beam in each direction, as shown in the attached diagram. Figure 5 As shown. The boundary identification process is as follows: For each rotation step, a light spot within the imaging area is identified until the light spot disappears. The moonlight azimuth vector at the exact moment the light spot disappears is recorded. Heliostat rotation angle , recorded as .

[0057] (7) Pointing error calculation: The pointing error calculation process consists of two steps: correcting the parameters of the heliostat calibration model and calculating the pointing error.

[0058] 1) Model parameter correction: A three-dimensional plane is established with the camera coordinates as the center point and the vector pointing from the camera to the heliostat as the normal. According to Formula 2, the current heliostat correction model parameters are used to calculate the beam boundary. The intersection point of the corresponding reflection vector with the three-dimensional plane is obtained by projecting this intersection point onto the three-dimensional plane to obtain a two-dimensional projection point. Calculate the convex hull of all projected points, and denote the center of the convex hull as . . Convex hull center The corresponding points are treated as camera coordinates, and then the projected points are corrected using the camera coordinates. The three-dimensional coordinates are denoted as the azimuth vector of the heliostat pointing to the projection point. Then, a set of optimal heliostat correction model parameters is calculated using regression analysis to minimize the average angular deviation between the azimuth vector of the actual projection point and the azimuth vector of the theoretical projection point. Let the optimal heliostat correction model parameters be denoted as... The regression analysis minimizes the function shown in Formula 7.

[0059] ) (7).

[0060] 2) According to Formula 1, input the moonlight azimuth vector and camera azimuth vector at the current moment, and use the original correction model parameters. Calculate the rotation angle of the heliostat, and denot it as . Using the corrected calibration model parameters Calculate the rotation angle of the heliostat So, the heliostat is in The formula for calculating the pointing error is as follows:

[0061] (8).

[0062] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for measuring the pointing error of a heliostat using moonlight and a camera, characterized in that, The heliostat and a camera are included, the heliostat is rotated, and the reflected moonlight is guided to the camera; the actual rotation angle of the heliostat when the center of the reflected light beam is located at the position of the camera is detected; at the same time, the current heliostat correction model parameters are used, and the theoretical rotation angle of the heliostat is calculated according to the moonlight azimuth vector and the camera azimuth vector; the difference between the actual rotation angle and the theoretical rotation angle is the pointing error of the heliostat at the current position; The heliostat correction model defines the relationship among the moonlight azimuth vector, the rotation angle of the heliostat, and the camera azimuth vector; The actual rotation angle calculation steps are as follows: S1. Beam positioning is performed; S2. Identify the light spot, rotate the heliostat to constantly change the position of the reflected light beam, so that the light spot imaging in the camera constantly changes, record the maximum value of the light spot imaging, use the DBSCAN algorithm to gather the pixel maximum values at different positions into a light spot class and a background class, and take the average of the minimum value of the light spot class and the maximum value of the background class as the brightness threshold for distinguishing the background and the light spot; S3. Beam profile identification, rotate the heliostat to make the reflected light beam move in N directions with the camera azimuth vector as the center, and identify the boundary of the light beam in each direction; S4. Beam center calculation, calculate the convex hull of the light beam profile boundary, and take the center of the convex hull as the beam center; S5. Actual rotation angle calculation: according to the beam boundary, the heliostat correction model parameters are corrected, and according to the corrected heliostat correction model parameters, the rotation angle of the heliostat when reflecting the moonlight to the beam center is calculated as the actual rotation angle.

2. The method of measuring pointing error of heliostat using moonlight and camera as claimed in claim 1, wherein, The calculation of the moonlight azimuth vector is based on the geometric projection method: A rectangular coordinate system is established with an observation point on the earth as the origin, i.e. the heliostat position is the coordinate origin, the X axis points to the geographic east, the Y axis points to the geographic north, and the Z axis is perpendicular to the horizontal plane and points upward; according to the lunar coordinates and the solar coordinates, a three-dimensional unit vector pointing to the center point of the moon-illuminated area by the sun from the coordinate origin is calculated, which is called a moonlight azimuth vector and is denoted as ; Meanwhile, the ratio of the area of the region of the moon illuminated by the sun to the area of the largest visible region of the moon is calculated, referred to as the moonlight visible region ratio, denoted as .

3. The method of measuring pointing error of heliostat using moonlight and camera as claimed in claim 1, wherein, The current time is determined whether to allow the use of moonlight measurement of the pointing error of the heliostat, the time allowed to measure the work must meet the following three conditions: the solar elevation angle must be less than the threshold value ; Moon elevation angle greater than threshold ; Moonlight viewable area percentage Need to be greater than threshold .

4. The method of measuring pointing error of heliostat using moonlight and camera as claimed in claim 1, wherein, Before measurement, it is necessary to judge whether the moonlight is visible, and the specific method is as follows: Select n The method comprises the following steps: selecting a heliostat which has completed calibration, controlling the heliostat to reflect moonlight to a camera, and identifying a reflected light spot of each heliostat. When a maximum value of maximum brightness of all reflected light spots of the heliostats is greater than a given threshold value, it is indicated that the current moonlight is visible.

5. The method of measuring pointing error of heliostat using moonlight and camera as claimed in claim 1, wherein, The light beam positioning step is as follows: the imaging area of the heliostat on the camera image is calculated, the moonlight orientation vector and the camera orientation vector are given, the theoretical rotation angle of the heliostat is calculated according to the correction model ; the heliostat is controlled to rotate to , and if the moonlight spot can be recognized in the imaging area, the light beam positioning is ended.

6. The method of measuring pointing error of a heliostat using moonlight and a camera according to claim 1, wherein, If no spot is recognized in the imaging area, search the area with a certain step size Modify the helical trajectory of the heliostat according to the detected spot. If a spot is detected, but the maximum brightness is less than a given threshold, search the surrounding area with the spot as the center, then move to the direction with the maximum brightness, and repeat the search process until the maximum brightness of the spot is greater than the given threshold.

Images