Biaxial full-degree-of-freedom compact portable photometer and compensation method

Abstract

The invention discloses a double-axis full-degree-of-freedom compact portable photometer and a compensation method, and the method comprises the steps: obtaining the three-axis gyroscope data of the portable photometer, the longitude and latitude of an observation point and a preset ephemeris, carrying out the inclination leveling of a pitch angle and a direction angle through the three-axis gyroscope data based on a pedestal of the photometer, and obtaining a compensation result. According to the longitude and latitude of an observation point, north pointing is carried out through a preset ephemeris, a celestial body is screened according to sky radiance distribution data after north pointing, biaxial wear optimization is carried out based on a celestial body edge pixel set and a current optical probe attitude angle, biaxial compensation parameters are obtained, and a light compensation result is obtained based on the biaxial compensation parameters. According to the method, a three-axis electronic gyroscope is used for automatically monitoring a double-degree-of-freedom comprehensive deviation generated during north pointing and light focusing adjustment and compensation installation of an instrument on line, the use difficulty is reduced, conductive slip rings are adopted in the pitching direction and the azimuth direction, the winding condition is avoided, and the rotational degree of freedom in the double-axis direction is improved.

Description

Technical Field

[0001] This invention relates to the field of photometers, and more particularly to a compact portable photometer with biaxial full degrees of freedom and a compensation method thereof. Background Technology

[0002] As an important instrument for measuring and monitoring atmospheric aerosols, the photometer is mainly used to measure the visible and near-infrared radiance of the sun, moon, and sky at different times and in different wavelengths. Before the initial installation and fixation, the photometer generally requires precise north-pointing and leveling of the mounting base. Moreover, during the initial installation and operation, the optical probe also needs to be aligned, which requires adjusting the initial azimuth position of the equipment and the initial pitch position of the optical probe. The initial alignment of the optical probe needs to be done during the daytime and in good weather conditions until the position of the light spot formed by sunlight passing through the aperture can be clearly identified. The north-pointing operation, base leveling, and optical probe alignment during installation increase the difficulty of using the instrument.

[0003] Existing photometers generally include modules such as an optical probe, a two-dimensional pointing turntable, and a control box. Because the two-dimensional pointing turntable carries the optical probe and moves in both pitch and azimuth directions, there is relative motion between the optical probe and other components. Therefore, communication between the optical probe and the control box, as well as between the control box and the two-dimensional pointing turntable, is often achieved through cable connections. During equipment installation, cable pulling is prone to occur. During the operation of the two-dimensional pointing turntable, it rotates continuously in both azimuth and pitch directions, causing the connecting cable between the optical head and the control box, or between the control box and the two-dimensional pointing turntable, to be in a state of constant dynamic stretching or bending. This can easily lead to photometer malfunctions and adversely affect the continuity of measurement data at unattended measurement sites. Summary of the Invention

[0004] The purpose of this invention is to provide a compact portable photometer with biaxial full degrees of freedom and a compensation method thereon.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution: This invention provides a compact portable photometer with biaxial full degrees of freedom and a compensation method, comprising: S1. Acquire the three-axis gyroscope data of the portable photometer, the latitude and longitude of the observation point, and the preset ephemeris. Based on the base of the photometer, adjust the pitch angle and azimuth angle by means of the three-axis gyroscope data. S2. Calculate the north alignment error using a preset ephemeris table based on the leveling results, system time, and latitude and longitude of the observation point. Then, align the optical probe of the photometer to the north based on the north alignment error to obtain the sky radiance distribution data after north alignment. S3, based on the sky radiance distribution data, celestial bodies are filtered by the direction of radiance change in the radiance abrupt region to obtain a set of celestial body edge pixels, wherein the celestial bodies include the sun and the moon; S4. Based on the celestial edge pixel set and the current optical probe attitude angle, the light adjustment parameters are decomposed by Lie group. Based on the light adjustment parameters, the dual-axis wear optimization is performed by Bayesian optimization to obtain the dual-axis compensation parameters. S5. Based on the dual-axis compensation parameters, perform zero-bias pre-estimation of the optical probe attitude angle, and based on the estimated zero-bias correction of the optical probe attitude angle, obtain the light compensation result.

[0006] Furthermore, the method for obtaining the leveling result includes: A fixed-duration sample sequence is generated based on the data from the three-axis gyroscope. The Allan variance of the sample sequence is calculated in a double logarithmic coordinate system. If the rate of change of the Allan variance is less than 5 degrees per hour, the photometer base is determined to be in a high-stability state. The three-axis gyroscope data includes roll angle, pitch angle, azimuth angle and corresponding angular velocity. The tilt deviation is obtained by performing difference calculations between the roll angle and pitch angle and the target horizontal attitude component. If the tilt deviation is less than 0.1 degrees, the preset photometer is determined to be leveled. If the tilt deviation is greater than or equal to 0.1 degrees, the roll direction and pitch direction are compensated in reverse by the dual-axis electric leveling feet of the photometer based on the tilt deviation, and the tilt deviation during the reverse compensation period is continuously calculated. The tilt deviation includes roll deviation and pitch deviation. If the tilt deviation is less than 0.1 degrees for 10 consecutive samplings, the leveling is completed. The leveling motor position is locked and the current gyroscope zero bias reference value is recorded to obtain the leveling result. The reverse compensation adjusts the motor speed by using a proportional-integral control algorithm. The proportional coefficient of the proportional-integral control algorithm is positively correlated with the absolute value of the current deviation.

[0007] Furthermore, the method for obtaining the sky radiance distribution data includes: Based on the photometer, the system time and the latitude and longitude of the observation point are obtained. The system time is synchronized by GPS positioning signal. The apparent right ascension and apparent declination of the celestial body at the current moment are calculated by linear interpolation based on the latitude and longitude of the observation point, the time synchronization, and the Chebyshev polynomial coefficients of the preset ephemeris. The local sidereal time is also calculated. The theoretical azimuth and theoretical altitude of the celestial body are obtained based on the local sidereal time. Based on the leveling result of the photometer, the current yaw angle of the optical probe is obtained through the zero bias reference value of the gyroscope. The difference between the current yaw angle and the theoretical azimuth angle is calculated to obtain the north alignment error. Based on the north alignment error, the yaw rotation is performed through the two-dimensional steering platform of the photometer until the north alignment error is less than 0.2 degrees and lasts for 0.5 seconds, then the north alignment is completed. Based on the optical probe, visible and near-infrared radiance data of the sky are acquired through automatic gain mode. If the saturation level of the visible and near-infrared radiance data is saturated, the exposure time of the automatic gain mode is halved. If the signal-to-noise ratio of the visible and near-infrared radiance data is less than 10dB, the exposure time is doubled. The automatic gain range is adjustable from 1 to 30dB, and sky radiance distribution data including the original gray value, azimuth angle and pitch angle are obtained.

[0008] Furthermore, the method for obtaining the set of pixels at the edge of the celestial body includes: Based on the sky radiance distribution data, the radiance values ​​of the scanning points are arranged in order of azimuth and elevation angles to obtain a spherical grid. The gradient of the sky radiance distribution data is calculated based on the spherical grid to obtain the radiance variation amplitude and direction of the pixels. The area with the highest pixel proportion in the sky radiance distribution data is taken as the background area. A dynamic detection threshold is generated based on the mean and standard deviation of the radiance of the background region. Pixels whose radiance changes exceed the dynamic detection threshold and whose change direction consistency is higher than 85% are marked as candidate mutation points. Candidate mutation points are then clustered into candidate mutation regions through eight-connected regions. The dispersion of the radiance change direction is calculated based on the candidate mutation region pixels. Regions with a dispersion of less than 15 degrees are retained. The roundness index of the candidate mutation region pixels is calculated based on spherical grid projection. Celestial regions with a roundness index in the range of 0.75 to 0.90 are retained. A circular outline is fitted to the celestial region using the centroid method, and the edge pixels of the circular outline are extracted to obtain a set of celestial edge pixels containing pixel coordinates, radiance change amplitude, and change direction.

[0009] Furthermore, the method for obtaining the biaxial compensation parameters includes: The pixel offset between the centroid coordinates of the celestial body and the field of view center of the optical probe is calculated based on the pixel set of the celestial body edge. The pixel offset is decomposed into angular offset according to the field of view angle and pixel resolution of the optical probe to obtain the attitude angle adjustment parameter. The current attitude angle of the optical probe is obtained according to the three-axis gyroscope data. Based on the attitude angle, the pixel coordinates of the optical probe field of view center are represented as a rotation matrix through Lie group decomposition to obtain the actual attitude matrix. Based on the actual attitude matrix and the attitude angle adjustment parameter, the target attitude matrix of the optical probe is obtained by mapping through Euler angle rotation matrix. The rotation range of the photometer's slip ring is divided into equal-angle intervals, and an indicative vector for each angle interval is generated. An angle interval weight vector is initialized based on the slip ring's historical usage time. The absolute difference between adjacent angles within the historical usage time series is calculated and summed for each equal-angle interval to obtain the total variation for each interval. The total variation of the slip ring angle is obtained by weighting the total variation and the angle interval weight vector. Based on the total variation, the angle interval weight vector, the actual attitude matrix, and the target attitude matrix, a slip ring rotation optimization objective function is constructed. The slip ring rotation optimization objective function is:

[0010] in Optimize the objective function for slip ring rotation. For the first The target pose rotation matrix for each sampling point For the first The actual attitude rotation matrix of each sampling point It is the Frobenius norm. This is the initial value of the wear avoidance coefficient. Wear weight, For the first The transpose of the corner interval weight vector of each sampling point This represents the total variation in the slip ring rotation angle; Based on the objective function of slip ring rotation optimization, Bayesian optimization is performed on the attitude angle adjustment parameters. If the total variation of slip ring rotation angle is reduced by more than or equal to 10% compared with before optimization, the corresponding attitude angle adjustment parameters are used as dual-axis compensation parameters. After each celestial tracking cycle, the rotation angle interval weight vector is updated by the least mean square algorithm according to the actual change of slip ring rotation angle.

[0011] Furthermore, the method for obtaining the corner interval weight vector includes: The formula for updating the weight of the turning interval is:

[0012] in For the first The corner interval weight vector of each sampling point For the first The corner interval indicative vector of each sampling point The learning rate for the least mean square algorithm is... As a preset small constant, we take 10. -6 , No. The change in slip ring angle at time t. No. The change in slip ring angle at time t. For the first Effective indication of celestial body tracking at each sampling point.

[0013] Furthermore, the method for obtaining the light compensation result includes: The zero-bias pre-estimation of the optical probe attitude angle is performed based on the dual-axis compensation parameters. The zero-bias pre-estimation formula is as follows:

[0014] in It is a set of pixels representing the edge of a celestial body. These are the pixel coordinates of the celestial body's edge. This is a sign function; if the input is positive, the output is 1; if the input is negative, the output is -1; if the input is zero, the output is 0. This represents the gradient direction deviation of the edge pixels. The gradient magnitude of the edge pixels. For time-varying observation matrices, For the first The optical probe azimuth angle of the frame, For the first The frame's optical probe pitch angle, The focal length of the optical probe. The coordinates of the principal point at the center of the field of view are obtained from the pixel coordinates corresponding to the image center. For the first The frame's gyroscope zero-bias vector includes roll zero-bias, pitch zero-bias, and azimuth zero-bias. To observe noise; The pre-estimated zero-bias vector is subtracted from the dual-axis compensation parameters to obtain the zero-bias drift elimination attitude angle. The pitch and azimuth axes of the photometer are then adjusted according to the zero-bias drift elimination attitude angle to obtain the light compensation result.

[0015] Furthermore, a compact portable photometer with biaxial full degrees of freedom is characterized by comprising: A two-dimensional pointing turntable and a base, wherein the two-dimensional pointing turntable includes a fixing component 4, a pitch direction conductive slip ring 5, a pitch direction rotation shaft 6, an azimuth direction conductive slip ring 7, an azimuth direction rotation shaft 9, and a three-axis electronic gyroscope 8. The optical probe 1 is mounted on the fixing member 4, the fixing member 4 is fixedly connected to the pitch direction rotation shaft 6, the pitch direction conductive slip ring 5 is installed at the central position of the end of the pitch direction rotation shaft 6, and the pitch direction conductive slip ring 5 is located inside the fixing member 4. The azimuth direction conductive slip ring 7 is installed at the end axis of the azimuth direction rotation shaft 9, the three-axis electronic gyroscope 8 is installed in the left housing on the upper part of the two-dimensional turntable, and the two-dimensional pointing turntable is rotatably connected to the base through the azimuth direction conductive slip ring 7; The optical probe 1 is electrically connected to the cable socket 2 via the cable 3. The cable socket 2, the pitch direction conductive slip ring 5, the azimuth direction conductive slip ring 7, and the three-axis electronic gyroscope 8 are electrically connected via the internal cable 10.

[0016] Furthermore, a compact portable photometer with dual-axis full degrees of freedom is characterized in that the pitch direction rotation axis 6 has no mechanical limit and the rotational degree of freedom is 0~360 degrees. When the pitch direction rotation axis 6 rotates, it drives the fixing part 4, optical probe 1, cable socket 2, cable 3, etc. to rotate as a whole. The fixing part 4, optical probe 1, cable socket 2, and cable 3 remain relatively stationary relative to each other. The azimuth rotation shaft 9 has no mechanical limit and has a rotational freedom of 0~360 degrees. When the azimuth rotation shaft 9 rotates, it drives the two-dimensional pointing turntable to rotate as a whole. The two-dimensional pointing turntable, the fixing part 4, the optical probe 1, and the cable 3 remain relatively stationary to each other.

[0017] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: This invention uses a three-axis electronic gyroscope to monitor the north orientation of the instrument online, automatically compensating for pointing and horizontal tilt errors during operation, thus improving the instrument's usability. By introducing a conductive slip ring into the cable, the cable and instrument components remain relatively stationary during rotation, reducing dynamic bending of the cable and exposure to the external environment. Optimizing the slip ring rotation partitions during compensation results in more uniform wear across the slip ring partitions, reducing the total variation in slip ring angle and the number of compensation steps, thereby extending the service life of the cable and conductive slip ring. By pre-calculating the zero bias caused by attitude angle changes in the compensation parameters, the comprehensive deviation of the motor rotation in two degrees of freedom during the compensation process is reduced, improving the continuity of the measurement process and measurement data. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the steps of a compact portable photometer with biaxial full degrees of freedom and a compensation method according to an embodiment of the present invention. Figure 2 This is a structural diagram of a compact portable photometer with biaxial full degrees of freedom according to an embodiment of the present invention; In the diagram: 1-Optical probe; 2-Cable socket; 3-Cable; 4-Fixed component; 5-Pitch direction conductive slip ring; 6-Pitch direction rotation axis; 7-Azimuth direction conductive slip ring; 8-Three-axis electronic gyroscope; 9-Azimuth direction rotation axis; 10-Internal cable. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] Reference Figure 1 As shown, the present invention provides a compact portable photometer with biaxial full degrees of freedom and a compensation method, comprising: S1. Acquire the three-axis gyroscope data of the portable photometer, the latitude and longitude of the observation point, and the preset ephemeris. Based on the base of the photometer, adjust the pitch angle and azimuth angle by means of the three-axis gyroscope data. In the actual evaluation, after the photometer was started, three-axis gyroscope data was continuously collected for 30 seconds and then filtered using a moving average. The fluctuation range of the filtered data was ≤0.02°. The Allan variance in logarithmic coordinates was calculated for the filtered roll and pitch angle data. The horizontal axis of the weighted logarithmic coordinates represents the sampling time interval, and the vertical axis represents the Allan variance fitting curve. The ratio of the difference between adjacent variances to the Allan variance was less than 5° / hour, and the variance converged to the stable interval. Specifically, the roll variance was ≤0.0004 and the pitch variance was ≤0.0003. The base was determined to have entered a high-stability state suitable for leveling. Taking the horizontal attitude of the photometer base as the target, with a theoretical target value of 0° roll and 0° pitch, the deviation between the current attitude and the target attitude was calculated. The roll deviation was 0.16° and the pitch deviation was 0.13°, both greater than the threshold of 0.1. °, activate 3 sets of electric leveling feet, adopt proportional-integral control algorithm, adjust according to the deviation reverse compensation principle, where the proportional coefficient is positively correlated with the absolute value of the deviation. When the deviation is >0.15°, the proportional coefficient is 2.5, when the deviation is 0.1°~0.15°, the proportional coefficient is 1.8, and when the deviation is <0.1°, the proportional coefficient is 1.2 to avoid leveling overshoot. The integral coefficient is fixed at 0.3. First adjust the pitch angle, then adjust the roll angle. After each round of adjustment, pause for 0.5 seconds, collect attitude data for feedback calibration, and continuously sample 10 times at 100ms intervals. The roll deviation and pitch deviation are ≤0.08° for a total of 1 second. Control the electric leveling feet to lock the position, with a self-locking torque ≥5N・m. At the same time, record the current gyroscope zero bias reference value, where the roll zero bias is 0.021°, the pitch zero bias is 0.012°, and the azimuth zero bias is 0.033°. S2. Calculate the north alignment error using a preset ephemeris table based on the leveling results, system time, and latitude and longitude of the observation point. Then, align the optical probe of the photometer to the north based on the north alignment error to obtain the sky radiance distribution data after north alignment. In the actual assessment, the system time was obtained and synchronized to standard time 02:30:00 via the GPS module. The Chebyshev polynomial coefficients from the preset ephemeris were then used to linearly interpolate the GMT time based on the solar orbit parameters for September 2024, yielding the apparent right ascension of the Sun at 19.75°, apparent declination at 5.2°, and local sidereal time at 18. h 55 m 43 s The hour angle is 264.17°. Based on the hour angle and the spherical trigonometric formula, the theoretical azimuth angle of the sun is calculated to be 180.0° and the theoretical altitude angle is 38.5°. Based on the zero bias reference value of the leveled gyroscope, the yaw angle output by the current gyroscope is read as 182.3°, so the north-alignment error is 2.3°. The azimuth axis of the two-dimensional steering platform is controlled to rotate at a speed of 0.5° / second, and the yaw angle feedback is collected every 0.1 seconds. When the north-alignment error is ≤0.18° after 5 consecutive samplings, the rotation is stopped, and the north-alignment is completed. Based on the optical probe after north alignment, the visible and near-infrared radiance data of the sky is collected. Its default exposure time is 10ms and the gain is 15dB. The first frame of sky radiance data is collected. The gray value of the first frame of data is 254. The exposure time is adjusted to 5ms. The data is continuously collected with an exposure time of 5ms and a gain of 15dB to obtain 100 frames of sky radiance distribution data containing the original gray values. S3, based on the sky radiance distribution data, celestial bodies are filtered by the direction of radiance change in the radiance abrupt region to obtain a set of celestial body edge pixels, wherein the celestial bodies include the sun and the moon; In the actual evaluation, each frame of the sky radiance distribution data was mapped to a spherical grid according to the pixel coordinate-attitude angle correspondence. The azimuth angle range was 175°~185°, and the pitch angle range was 35°~42°. The grid resolution was consistent with the image pixels, with one pixel corresponding to approximately 0.0078° of attitude angle. For each pixel in the spherical grid, a 3×3 Sobel operator was used to calculate the radiance gradient, obtaining the gradient magnitude and gradient direction. The gradient magnitude represents the intensity of the radiance change at the pixel, including horizontal and vertical gradient components. The gradient direction is the angle along the direction of increasing radiance, ranging from 0° to 360°. A 100×100 pixel area in the upper left corner of the image was selected as the background area, free from sun and cloud interference, with uniform radiance. The mean of the background gradient magnitude was 5, and the standard deviation was 2, providing a benchmark for subsequent threshold setting. The edge candidate threshold is calculated using a dynamic thresholding rule that adds 3 times the standard deviation of the background mean. The edge candidate threshold is 11. The entire frame of pixels is traversed, and candidate mutation points that meet the requirements of gradient magnitude greater than the edge candidate threshold and gradient direction consistency ≥ 85% are marked. Gradient direction consistency is calculated by using a 3×3 neighborhood as a unit and counting the percentage of pixels whose gradient direction deviation from the center pixel direction is ≤ 10°. Isolated noise points are excluded, resulting in 128 marked candidate mutation points for a single frame image. Eight-connected component analysis is performed on the candidate mutation points, and finally, two candidate mutation regions are obtained by clustering. Region 1 contains 112 pixels, and Region 2 is an isolated noise region of clouds containing 16 pixels. The standard deviation of the gradient direction of all pixels in the region is calculated. The dispersion of Region 1 is 11°, and the dispersion of Region 2 is 42°. Region 2 is removed. The roundness of Region 1 is 0.85, and this region is retained as the target edge region. The retained target edge region is fitted with a circular contour using the centroid method. Its centroid coordinates are (642, 365), which corresponds to 2 pixels to the right and 5 pixels above the image center. The calculated contour radius is 7 pixels with the centroid as the center. All pixels on the circular contour are extracted to form a set of 44 pixels of the sun edge, including the centroid pixel coordinates, gradient magnitude and gradient direction. The above process is repeated for 100 frames of data. 3 frames with slight cloud interference are removed to obtain a set of 97 effective edge pixels. S4. Based on the celestial edge pixel set and the current optical probe attitude angle, the light adjustment parameters are decomposed by Lie group. Based on the light adjustment parameters, the dual-axis wear optimization is performed by Bayesian optimization to obtain the dual-axis compensation parameters. In the actual evaluation, the pixel offset between the centroid coordinates of the celestial body and the center of the optical probe's field of view is calculated based on the pixel set at the edge of the celestial body. The centroid pixel coordinates are (642, 365), and the center of the optical probe's field of view is (640, 360). The pixel offsets are 2 pixels and 5 pixels, respectively. Based on the field of view and pixel resolution (128 pixels / ° horizontally and 128 pixels / ° vertically), the pixel offset is converted into an angle offset. The azimuth angle is adjusted to the left by -0.015625°, and the pitch angle is adjusted upward by 0.0390625° to obtain the attitude angle adjustment parameters. Based on the current attitude angle of the optical probe (azimuth angle 180.0°, pitch angle 38.5°, roll angle 0°), it is represented as SO(3) rotation matrix through Lie group decomposition to obtain the actual attitude matrix. The attitude angle is then adjusted. The parameters are superimposed on the current attitude angle, and the target attitude angle (179.984375°, 38.5390625°) is obtained through Euler angle rotation matrix mapping. The conductive slip rings installed in the two fully free directions of the portable photometer are obtained, and the rotation range of the slip rings is divided into 12 equal-angle intervals, numbered 0~11. The current azimuth angle of 180.0° falls in interval 6. The rotation interval weight vector is initialized according to the historical usage time of the slip rings as [1,1,1,1,1,1,2,1,1,1,1, [1,1], where interval 6 was used for a cumulative total of 12 hours, and the other intervals averaged 60 hours. The pitch angle of 38.53° fell in interval 1 and had a weight of 1. For the rotation time series within a tracking period of 100 frames, the sum of the absolute differences between adjacent rotation angles was calculated by segmenting them into equal angle intervals. The total segmented variation of azimuth angle interval 6 was 5.0°, and the total segmented variation of the other 11 intervals was 1.0° each. Combined with the weight vector, a weighted sum of 21° was obtained. Based on the slip ring rotation optimization objective function, the attitude angle adjustment parameters were adjusted. Yeats optimization was performed, with the initial wear avoidance coefficient set to 0.1 and the wear weight set to 0.5, until the total variation of the slip ring rotation angle decreased to 18.9° and the pitch angle decreased to 0.9°, resulting in dual-axis compensation parameters of -0.015° and +0.039°. After the tracking cycle ended, the weight vector was updated using the least mean square algorithm based on the actual change in slip ring rotation angle, resulting in the weight of interval 6 being updated from 2 to 2.05, while the weights of other intervals remained unchanged. The LMS learning rate was 0.1, and the celestial tracking was effective. S5. Based on the dual-axis compensation parameters, perform zero-bias pre-estimation of the optical probe attitude angle, and based on the estimated zero-bias correction of the optical probe attitude angle, obtain the light compensation result.

[0021] In practical evaluation, based on the intrinsic parameters of the optical probe and the dual-axis compensation parameters, a 3×3 matrix of time-varying observation matrix is ​​constructed according to the pinhole camera projection model, and updated in real time with the attitude angle. The focal length is 8mm, the principal point coordinates are (640, 360), and the dual-axis compensation parameter is 179.98°. It is -0.00028. The initial value is -1.0. The average coordinates of the edge pixels in the first frame are (642, 365). The gradient direction deviation and gradient magnitude of each pixel are extracted from the set of celestial edge pixels. The theoretical tangent direction at this position on the solar edge is 270°±5° along the radial perpendicular direction. The deviation range of a single pixel is -5° to +5°. 44 observation sample values ​​are obtained by weighting the sign function and magnitude on the left side of the zero-bias pre-estimation formula. The range of these values ​​is -28 to +25. The average of the 44 samples is taken, and the left-side observation value is 0.82. The observation value and the time-varying observation matrix are substituted into the zero-bias pre-estimation formula, and the overdetermined equations are solved by the least squares method to obtain the real-time zero-bias vector [0.019°, 0.010°, 0.031°]T and the initial zero bias. For benchmark comparison, the zero-bias drift amounts were -0.002° roll, -0.002° pitch, and -0.002° azimuth. Zero-bias elimination correction was performed using the pre-estimated dual-axis compensation parameters of the zero-bias vector. The corrected attitude angles were 179.953° azimuth, 38.529° pitch, and -0.019° roll. Based on the zero-bias attitude angles, the pitch and azimuth axes of the photometer were adjusted. The compensation results were sent to the dual-axis turntable controller, which controlled the pitch and azimuth rotation axes to be fine-tuned according to the correction angles. The fine-tuning amounts were -0.031° and -0.010°, respectively. After fine-tuning, 10 frames of attitude data were continuously acquired, with the attitude angle fluctuation ≤0.005°, completing the entire light compensation process.

[0022] In this embodiment, the method for obtaining the leveling result includes: A fixed-duration sample sequence is generated based on the data from the three-axis gyroscope. The Allan variance of the sample sequence is calculated in a double logarithmic coordinate system. If the rate of change of the Allan variance is less than 5 degrees per hour, the photometer base is determined to be in a high-stability state. The three-axis gyroscope data includes roll angle, pitch angle, azimuth angle and corresponding angular velocity. The tilt deviation is obtained by performing difference calculations between the roll angle and pitch angle and the target horizontal attitude component. If the tilt deviation is less than 0.1 degrees, the preset photometer is determined to be leveled. If the tilt deviation is greater than or equal to 0.1 degrees, the roll direction and pitch direction are compensated in reverse by the dual-axis electric leveling feet of the photometer based on the tilt deviation, and the tilt deviation during the reverse compensation period is continuously calculated. The tilt deviation includes roll deviation and pitch deviation. If the tilt deviation is less than 0.1 degrees for 10 consecutive samplings, the leveling is completed. The leveling motor position is locked and the current gyroscope zero bias reference value is recorded to obtain the leveling result. The reverse compensation adjusts the motor speed by using a proportional-integral control algorithm. The proportional coefficient of the proportional-integral control algorithm is positively correlated with the absolute value of the current deviation.

[0023] In this embodiment, the method for obtaining the sky radiance distribution data includes: Based on the photometer, the system time and the latitude and longitude of the observation point are obtained. The system time is synchronized by GPS positioning signal. The apparent right ascension and apparent declination of the celestial body at the current moment are calculated by linear interpolation based on the latitude and longitude of the observation point, the time synchronization, and the Chebyshev polynomial coefficients of the preset ephemeris. The local sidereal time is also calculated. The theoretical azimuth and theoretical altitude of the celestial body are obtained based on the local sidereal time. Based on the leveling result of the photometer, the current yaw angle of the optical probe is obtained through the zero bias reference value of the gyroscope. The difference between the current yaw angle and the theoretical azimuth angle is calculated to obtain the north alignment error. Based on the north alignment error, the yaw rotation is performed through the two-dimensional steering platform of the photometer until the north alignment error is less than 0.2 degrees and lasts for 0.5 seconds, then the north alignment is completed. Based on the optical probe, visible and near-infrared radiance data of the sky are acquired through automatic gain mode. If the saturation level of the visible and near-infrared radiance data is saturated, the exposure time of the automatic gain mode is halved. If the signal-to-noise ratio of the visible and near-infrared radiance data is less than 10dB, the exposure time is doubled. The automatic gain range is adjustable from 1 to 30dB, and sky radiance distribution data including the original gray value, azimuth angle and pitch angle are obtained.

[0024] In this embodiment, the method for obtaining the set of celestial body edge pixels includes: Based on the sky radiance distribution data, the radiance values ​​of the scanning points are arranged in order of azimuth and elevation angles to obtain a spherical grid. The gradient of the sky radiance distribution data is calculated based on the spherical grid to obtain the radiance variation amplitude and direction of the pixels. The area with the highest pixel proportion in the sky radiance distribution data is taken as the background area. A dynamic detection threshold is generated based on the mean and standard deviation of the radiance of the background region. Pixels whose radiance changes exceed the dynamic detection threshold and whose change direction consistency is higher than 85% are marked as candidate mutation points. Candidate mutation points are then clustered into candidate mutation regions through eight-connected regions. The dispersion of the radiance change direction is calculated based on the candidate mutation region pixels. Regions with a dispersion of less than 15 degrees are retained. The roundness index of the candidate mutation region pixels is calculated based on spherical grid projection. Celestial regions with a roundness index in the range of 0.75 to 0.90 are retained. A circular outline is fitted to the celestial region using the centroid method, and the edge pixels of the circular outline are extracted to obtain a set of celestial edge pixels containing pixel coordinates, radiance change amplitude, and change direction.

[0025] In this embodiment, the method for obtaining the biaxial compensation parameters includes: The pixel offset between the centroid coordinates of the celestial body and the field of view center of the optical probe is calculated based on the pixel set of the celestial body edge. The pixel offset is decomposed into angular offset according to the field of view angle and pixel resolution of the optical probe to obtain the attitude angle adjustment parameter. The current attitude angle of the optical probe is obtained according to the three-axis gyroscope data. Based on the attitude angle, the pixel coordinates of the optical probe field of view center are represented as a rotation matrix through Lie group decomposition to obtain the actual attitude matrix. Based on the actual attitude matrix and the attitude angle adjustment parameter, the target attitude matrix of the optical probe is obtained by mapping through Euler angle rotation matrix. The rotation range of the photometer's slip ring is divided into equal-angle intervals, and an indicative vector for each angle interval is generated. An angle interval weight vector is initialized based on the slip ring's historical usage time. The absolute difference between adjacent angles within the historical usage time series is calculated and summed for each equal-angle interval to obtain the total variation for each interval. The total variation of the slip ring angle is obtained by weighting the total variation and the angle interval weight vector. Based on the total variation, the angle interval weight vector, the actual attitude matrix, and the target attitude matrix, a slip ring rotation optimization objective function is constructed. The slip ring rotation optimization objective function is:

[0026] in Optimize the objective function for slip ring rotation. For the first The target pose rotation matrix for each sampling point For the first The actual attitude rotation matrix of each sampling point It is the Frobenius norm. This is the initial value of the wear avoidance coefficient. Wear weight, For the first The transpose of the corner interval weight vector of each sampling point This represents the total variation in the slip ring rotation angle; Based on the objective function of slip ring rotation optimization, Bayesian optimization is performed on the attitude angle adjustment parameters. If the total variation of slip ring rotation angle is reduced by more than or equal to 10% compared with before optimization, the corresponding attitude angle adjustment parameters are used as dual-axis compensation parameters. After each celestial tracking cycle, the rotation angle interval weight vector is updated by the least mean square algorithm according to the actual change of slip ring rotation angle.

[0027] In this embodiment, the method for obtaining the corner interval weight vector includes: The formula for updating the weight of the turning interval is:

[0028] in For the first The corner interval weight vector of each sampling point For the first The corner interval indicative vector of each sampling point The learning rate for the least mean square algorithm is... As a preset small constant, we take 10. -6 , No. The change in slip ring angle at time t. No. The change in slip ring angle at time t. For the first Effective indication of celestial body tracking at each sampling point.

[0029] In this embodiment, the method for obtaining the light compensation result includes: The zero-bias pre-estimation of the optical probe attitude angle is performed based on the dual-axis compensation parameters. The zero-bias pre-estimation formula is as follows:

[0030] in It is a set of pixels representing the edge of a celestial body. These are the pixel coordinates of the celestial body's edge. This is a sign function; if the input is positive, the output is 1; if the input is negative, the output is -1; if the input is zero, the output is 0. This represents the gradient direction deviation of the edge pixels. The gradient magnitude of the edge pixels. For time-varying observation matrices, For the first The optical probe azimuth angle of the frame, For the first The frame's optical probe pitch angle, The focal length of the optical probe. The coordinates of the principal point at the center of the field of view are obtained from the pixel coordinates corresponding to the image center. For the first The frame's gyroscope zero-bias vector includes roll zero-bias, pitch zero-bias, and azimuth zero-bias. To observe noise; The pre-estimated zero-bias vector is subtracted from the dual-axis compensation parameters to obtain the zero-bias drift elimination attitude angle. The pitch and azimuth axes of the photometer are then adjusted according to the zero-bias drift elimination attitude angle to obtain the light compensation result.

[0031] In this embodiment, a compact portable photometer with biaxial full degrees of freedom is characterized by comprising: A two-dimensional pointing turntable and a base, wherein the two-dimensional pointing turntable includes a fixing component 4, a pitch direction conductive slip ring 5, a pitch direction rotation shaft 6, an azimuth direction conductive slip ring 7, an azimuth direction rotation shaft 9, and a three-axis electronic gyroscope 8. The optical probe 1 is mounted on the fixing member 4, the fixing member 4 is fixedly connected to the pitch direction rotation shaft 6, the pitch direction conductive slip ring 5 is installed at the central position of the end of the pitch direction rotation shaft 6, and the pitch direction conductive slip ring 5 is located inside the fixing member 4. The azimuth direction conductive slip ring 7 is installed at the end axis of the azimuth direction rotation shaft 9, the three-axis electronic gyroscope 8 is installed in the left housing on the upper part of the two-dimensional turntable, and the two-dimensional pointing turntable is rotatably connected to the base through the azimuth direction conductive slip ring 7; The optical probe 1 is electrically connected to the cable socket 2 via the cable 3. The cable socket 2, the pitch direction conductive slip ring 5, the azimuth direction conductive slip ring 7, and the three-axis electronic gyroscope 8 are electrically connected via the internal cable 10.

[0032] In this embodiment, a compact portable photometer with dual-axis full degrees of freedom is characterized in that the pitch direction rotation axis 6 has no mechanical limit and the rotational degree of freedom is 0~360 degrees. When the pitch direction rotation axis 6 rotates, it drives the fixing part 4, optical probe 1, cable socket 2, cable 3, etc. to rotate as a whole. The fixing part 4, optical probe 1, cable socket 2, and cable 3 remain relatively stationary relative to each other. The azimuth rotation shaft 9 has no mechanical limit and has a rotational freedom of 0~360 degrees. When the azimuth rotation shaft 9 rotates, it drives the two-dimensional pointing turntable to rotate as a whole. The two-dimensional pointing turntable, the fixing part 4, the optical probe 1, and the cable 3 remain relatively stationary to each other.

[0033] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

Claims

1. A compact portable photometer with biaxial full degrees of freedom and a compensation method thereof, characterized in that, Includes the following steps: S1. Acquire the three-axis gyroscope data of the portable photometer, the latitude and longitude of the observation point, and the preset ephemeris. Based on the base of the photometer, adjust the pitch angle and azimuth angle by means of the three-axis gyroscope data. S2. Calculate the north alignment error using a preset ephemeris table based on the leveling results, system time, and latitude and longitude of the observation point. Then, align the optical probe of the photometer to the north based on the north alignment error to obtain the sky radiance distribution data after north alignment. S3, based on the sky radiance distribution data, celestial bodies are filtered by the direction of radiance change in the radiance abrupt region to obtain a set of celestial body edge pixels, wherein the celestial bodies include the sun and the moon; S4. Based on the celestial edge pixel set and the current optical probe attitude angle, the light adjustment parameters are decomposed by Lie group. Based on the light adjustment parameters, the dual-axis wear optimization is performed by Bayesian optimization to obtain the dual-axis compensation parameters. S5. Based on the dual-axis compensation parameters, perform zero-bias pre-estimation of the optical probe attitude angle, and based on the estimated zero-bias correction of the optical probe attitude angle, obtain the light compensation result.

2. The compact portable photometer and compensation method with biaxial full degrees of freedom according to claim 1, characterized in that, The method for obtaining the leveling result includes: A fixed-duration sample sequence is generated based on the data from the three-axis gyroscope. The Allan variance of the sample sequence is calculated in a double logarithmic coordinate system. If the rate of change of the Allan variance is less than 5 degrees per hour, the photometer base is determined to be in a high-stability state. The three-axis gyroscope data includes roll angle, pitch angle, azimuth angle and corresponding angular velocity. The tilt deviation is obtained by performing difference calculations between the roll angle and pitch angle and the target horizontal attitude component. If the tilt deviation is less than 0.1 degrees, the preset photometer is determined to be leveled. If the tilt deviation is greater than or equal to 0.1 degrees, the roll direction and pitch direction are compensated in reverse by the dual-axis electric leveling feet of the photometer based on the tilt deviation, and the tilt deviation during the reverse compensation period is continuously calculated. The tilt deviation includes roll deviation and pitch deviation. If the tilt deviation is less than 0.1 degrees for 10 consecutive samplings, the leveling is completed. The leveling motor position is locked and the current gyroscope zero bias reference value is recorded to obtain the leveling result. The reverse compensation adjusts the motor speed by using a proportional-integral control algorithm. The proportional coefficient of the proportional-integral control algorithm is positively correlated with the absolute value of the current deviation.

3. The compact portable photometer and compensation method with biaxial full degrees of freedom according to claim 1, characterized in that, The method for obtaining the sky radiance distribution data includes: Based on the photometer, the system time and the latitude and longitude of the observation point are obtained. The system time is synchronized by GPS positioning signal. The apparent right ascension and apparent declination of the celestial body at the current moment are calculated by linear interpolation based on the latitude and longitude of the observation point, the time synchronization, and the Chebyshev polynomial coefficients of the preset ephemeris. The local sidereal time is also calculated. The theoretical azimuth and theoretical altitude of the celestial body are obtained based on the local sidereal time. Based on the leveling result of the photometer, the current yaw angle of the optical probe is obtained through the zero bias reference value of the gyroscope. The difference between the current yaw angle and the theoretical azimuth angle is calculated to obtain the north alignment error. Based on the north alignment error, the yaw rotation is performed through the two-dimensional steering platform of the photometer until the north alignment error is less than 0.2 degrees and lasts for 0.5 seconds, then the north alignment is completed. Based on the optical probe, visible and near-infrared radiance data of the sky are acquired through automatic gain mode. If the saturation level of the visible and near-infrared radiance data is saturated, the exposure time of the automatic gain mode is halved. If the signal-to-noise ratio of the visible and near-infrared radiance data is less than 10dB, the exposure time is doubled. The automatic gain range is adjustable from 1 to 30dB, and sky radiance distribution data including the original gray value, azimuth angle and pitch angle are obtained.

4. The compact portable photometer and compensation method with biaxial full degrees of freedom according to claim 1, characterized in that, The method for obtaining the set of pixels at the edge of the celestial body includes: Based on the sky radiance distribution data, the radiance values ​​of the scanning points are arranged in order of azimuth and elevation angles to obtain a spherical grid. The gradient of the sky radiance distribution data is calculated based on the spherical grid to obtain the radiance variation amplitude and direction of the pixels. The area with the highest pixel proportion in the sky radiance distribution data is taken as the background area. A dynamic detection threshold is generated based on the mean and standard deviation of the radiance of the background region. Pixels whose radiance changes exceed the dynamic detection threshold and whose change direction consistency is higher than 85% are marked as candidate mutation points. Candidate mutation points are then clustered into candidate mutation regions through eight-connected regions. The dispersion of the radiance change direction is calculated based on the candidate mutation region pixels. Regions with a dispersion of less than 15 degrees are retained. The roundness index of the candidate mutation region pixels is calculated based on spherical grid projection. Celestial regions with a roundness index in the range of 0.75 to 0.90 are retained. A circular outline is fitted to the celestial region using the centroid method, and the edge pixels of the circular outline are extracted to obtain a set of celestial edge pixels containing pixel coordinates, radiance change amplitude, and change direction.

5. The compact portable photometer and compensation method with biaxial full degrees of freedom according to claim 1, characterized in that, The method for obtaining the biaxial compensation parameters includes: The pixel offset between the centroid coordinates of the celestial body and the field of view center of the optical probe is calculated based on the pixel set of the celestial body edge. The pixel offset is decomposed into angular offset according to the field of view angle and pixel resolution of the optical probe to obtain the attitude angle adjustment parameter. The current attitude angle of the optical probe is obtained according to the three-axis gyroscope data. Based on the attitude angle, the pixel coordinates of the optical probe field of view center are represented as a rotation matrix through Lie group decomposition to obtain the actual attitude matrix. Based on the actual attitude matrix and the attitude angle adjustment parameter, the target attitude matrix of the optical probe is obtained by mapping through Euler angle rotation matrix. The rotation range of the photometer's slip ring is divided into equal-angle intervals, and an indicative vector for each angle interval is generated. An angle interval weight vector is initialized based on the slip ring's historical usage time. The absolute difference between adjacent angles within the historical usage time series is calculated and summed for each equal-angle interval to obtain the total variation for each interval. The total variation of the slip ring angle is obtained by weighting the total variation and the angle interval weight vector. Based on the total variation, the angle interval weight vector, the actual attitude matrix, and the target attitude matrix, a slip ring rotation optimization objective function is constructed. The slip ring rotation optimization objective function is: ； in Optimize the objective function for slip ring rotation. For the first The target pose rotation matrix for each sampling point For the first The actual attitude rotation matrix of each sampling point It is the Frobenius norm. This is the initial value of the wear avoidance coefficient. Wear weight, For the first The transpose of the corner interval weight vector of each sampling point This represents the total variation in the slip ring rotation angle; Based on the objective function of slip ring rotation optimization, Bayesian optimization is performed on the attitude angle adjustment parameters. If the total variation of slip ring rotation angle is reduced by more than or equal to 10% compared with before optimization, the corresponding attitude angle adjustment parameters are used as dual-axis compensation parameters. After each celestial tracking cycle, the rotation angle interval weight vector is updated by the least mean square algorithm according to the actual change of slip ring rotation angle.

6. The compact portable photometer and compensation method with biaxial full degrees of freedom according to claim 1, characterized in that, The method for obtaining the corner interval weight vector includes: The formula for updating the weight of the turning interval is: ； in For the first The corner interval weight vector of each sampling point For the first The corner interval indicative vector of each sampling point The learning rate for the least mean square algorithm is... As a preset small constant, we take 10. -6 , No. The change in slip ring angle at time t. No. The change in slip ring angle at time t. For the first Effective indication of celestial body tracking at each sampling point.

7. A compact portable photometer with biaxial full degrees of freedom and a compensation method according to claim 1, characterized in that, The method for obtaining the light compensation result includes: The zero-bias pre-estimation of the optical probe attitude angle is performed based on the dual-axis compensation parameters. The zero-bias pre-estimation formula is as follows: ； in It is a set of pixels representing the edge of a celestial body. These are the pixel coordinates of the celestial body's edge. This is a sign function; if the input is positive, the output is 1; if the input is negative, the output is -1; if the input is zero, the output is 0. This represents the gradient direction deviation of the edge pixels. The gradient magnitude of the edge pixels. For time-varying observation matrices, For the first The optical probe azimuth angle of the frame, For the first The frame's optical probe pitch angle, The focal length of the optical probe. The coordinates of the principal point at the center of the field of view are obtained from the pixel coordinates corresponding to the image center. For the first The frame's gyroscope zero-bias vector includes roll zero-bias, pitch zero-bias, and azimuth zero-bias. To observe noise; The pre-estimated zero-bias vector is subtracted from the dual-axis compensation parameters to obtain the zero-bias drift elimination attitude angle. The pitch and azimuth axes of the photometer are then adjusted according to the zero-bias drift elimination attitude angle to obtain the light compensation result.

8. A compact portable photometer with biaxial full degrees of freedom according to claim 1, characterized in that, include: A two-dimensional pointing turntable and a base, wherein the two-dimensional pointing turntable includes a fixing component 4, a pitch direction conductive slip ring 5, a pitch direction rotation shaft 6, an azimuth direction conductive slip ring 7, an azimuth direction rotation shaft 9, and a three-axis electronic gyroscope 8. The optical probe 1 is mounted on the fixing member 4, the fixing member 4 is fixedly connected to the pitch direction rotation shaft 6, the pitch direction conductive slip ring 5 is installed at the central position of the end of the pitch direction rotation shaft 6, and the pitch direction conductive slip ring 5 is located inside the fixing member 4. The azimuth direction conductive slip ring 7 is installed at the end axis of the azimuth direction rotation shaft 9, the three-axis electronic gyroscope 8 is installed in the left housing on the upper part of the two-dimensional turntable, and the two-dimensional pointing turntable is rotatably connected to the base through the azimuth direction conductive slip ring 7; The optical probe 1 is electrically connected to the cable socket 2 via the cable 3. The cable socket 2, the pitch direction conductive slip ring 5, the azimuth direction conductive slip ring 7, and the three-axis electronic gyroscope 8 are electrically connected via the internal cable 10.

9. A compact portable photometer with biaxial full degrees of freedom according to claim 8, characterized in that, include: The pitch direction rotation shaft 6 has no mechanical limit and has a rotational freedom of 0~360 degrees. When the pitch direction rotation shaft 6 rotates, it drives the fixing part 4, optical probe 1, cable socket 2, cable 3, etc. to rotate as a whole. The fixing part 4, optical probe 1, cable socket 2, and cable 3 remain relatively stationary to each other. The azimuth rotation shaft 9 has no mechanical limit and has a rotational freedom of 0~360 degrees. When the azimuth rotation shaft 9 rotates, it drives the two-dimensional pointing turntable to rotate as a whole. The two-dimensional pointing turntable, the fixing part 4, the optical probe 1, and the cable 3 remain relatively stationary to each other.

Images