A method for measuring and compensating heliostat backlash error during mirror field operation

Abstract

The invention belongs to the technical field of heliostat control, and in particular relates to a method for measuring and compensating a backlash error of a heliostat during mirror field operation, comprising the following steps: S1. calculating a search area of ​​a heliostat on a camera image; S2. error sampling, calculating a backlash error, and evaluating the collected backlash error; S3. regression analysis, obtaining a relationship between the backlash error and a stepping position; S4. performing error compensation, and driving a motor to move to a designed target position after error compensation; the method has the advantage that the backlash error can be measured and compensated for each heliostat individually, and has higher precision.

Description

Technical Field

[0001] The invention belongs to the technical field of heliostat control, and in particular relates to a method for measuring and compensating a backlash error of a heliostat during operation of a mirror field. Background Art

[0002] The transmission device of the heliostat usually contains gears. When the gear changes its direction of rotation, due to the existence of transmission backlash, the gear will rotate, but the heliostat will not rotate immediately. This error introduced by the transmission backlash is called backlash error. This part of the error exists objectively and cannot be eliminated. Heliostat control is divided into two modes: open-loop control and closed-loop control. The closed-loop control technology relies on angle sensor devices to directly control the mirror angle and is not affected by the backlash error. However, the additional equipment increases the cost, and the positioning process requires multiple adjustments to the heliostat posture, which takes a long time. The open-loop control technology cannot detect the mirror angle, but calculates the theoretical angle of the mirror through models and algorithms. It can usually achieve high tracking accuracy and does not require additional equipment. It has the advantages of low cost and fast tracking, so it is widely used in commercial projects.

[0003] Since open-loop control technology cannot directly measure the mirror angle, it is necessary to model and compensate for the backlash error. Traditional open-loop control technology samples some heliostats to detect backlash in the laboratory, uses the statistical mean to represent the backlash error of all heliostats, and compensates when the heliostats move. However, the backlash of each heliostat is generally different, and the backlash will change as the heliostat wears during use, so the method of sampling and measuring to compensate for backlash error in the laboratory is less accurate. Summary of the invention

[0004] Based on the above problems, the present invention proposes a method for measuring and compensating the backlash error of a heliostat during the operation of the mirror field. The backlash error of the heliostat at several rotation angles is measured by sampling with a light source and a camera, and the backlash error of the heliostat at any rotation angle is calculated by a regression analysis method. Finally, the backlash error is compensated when the heliostat changes direction. Compared with the traditional backlash compensation method, it has the advantages of low cost, high precision and high efficiency. The technical solution is as follows:

[0005] A method for measuring and compensating a backlash error of a heliostat during operation of a mirror field comprises the following steps:

[0006] S1. Calculate the heliostat search area on the camera image;

[0007] S2. Error sampling, calculating backlash error, and evaluating the collected backlash error;

[0008] S3. Regression analysis to obtain the relationship between backlash error and step position;

[0009] S4. Perform error compensation and drive the motor to move to the designed target position after error compensation.

[0010] Preferably, in step S1, the search area calculation process is as follows: the heliostat is facing the camera, the position of the heliostat vertex on the image is calculated using the camera imaging model, and then the minimum outer expansion rectangle of all vertex pixel positions is calculated as the theoretical search area, and finally the set pixel range is expanded outward to obtain the final search area.

[0011] Preferably, in step S2, for a given three-element combination of light source, heliostat and camera, when the stepping position is in the reflection area, it means that the light source can be reflected into the camera, that is, the reflection area is a mapping of the reflection spot in the stepping position space, and the reflection area can measure four backlash errors. The single backlash error measurement process is as follows:

[0012] S21. Rotate the heliostat to the center of the reflection area;

[0013] S22. Fix the other drive axis positions at the set step interval Gradually reduce the current measured drive shaft step position, where E0 is the typical backlash error; take a photo to identify the light spot at each move, and calculate the light spot intensity , spot intensity is the sum of the grayscale values ​​of all light spot pixels, and the formula is as follows:

[0014] ;

[0015] Where n is the number of spot pixels, Gray k is the gray value of the kth spot pixel;

[0016] S23. Continue to drive the stepper motor to make the camera at the edge of the reflected light spot. At this time, the light spot disappears. Record this position as , the distance the motor moves from the camera being at the center of the light spot to the camera being at the edge of the light spot is recorded as L1;

[0017] S24. Continue to drive the stepper motor for a distance L2=4*E0; change the direction of the motor to δ e Move the distance L1+L2 at intervals and record the intensity of the light spot after each movement. ;

[0018] Backlash error at The definition is as follows:

[0019] ;

[0020] The spot intensity is a curve represented by several discrete data, and a continuous position is calculated by linear interpolation. Theoretical spot intensity at:

[0021] ;

[0022] in is less than The maximum known position of '' is greater than The minimum known position of

[0023] S25. Repeat the above steps to measure the backlash errors at other positions in the reflection area.

[0024] Preferably, the acquired backlash error is evaluated:

[0025] Calculate the relative error of two error samples for the same drive axis in the reflection area ,if If the evaluation is passed, the two error samples need to be discarded. The calculation formula is as follows:

[0026] ;

[0027] Repeat steps S21 to S25 for a single heliostat until enough error samples are collected. The regression equation of backlash error and step position determines the minimum number of samples, which is recorded as N. min .

[0028] Preferably, in step S3, the specific steps are as follows:

[0029] S31. The backlash error data corresponding to several step positions are obtained through the error sampling process. ;

[0030] S32. Select the regression equation according to the mechanical structure of the drive and the factors affecting the backlash error, and convert the backlash error The relationship between the step position x is recorded as ,in is a set of parameter vectors of the regression equation;

[0031] S33. Use the least squares method to solve and find the parameter vector The best estimate of , so that the sum of squared errors of the samples is minimized:

[0032] ;

[0033] S34. Calculate the coefficient of determination R 2 , evaluate the regression equation parameter vector Is it acceptable? 2 The calculation formula is as follows:

[0034] ;

[0035] S35. If R 2 >R 2 limit , then accept the vector Otherwise, remove the sample with the largest deviation and repeat steps S31 to S34. If the number of samples is less than or equal to N min , it means that the existing samples cannot complete the regression analysis and it is necessary to continue collecting samples.

[0036] Preferably, in step S4, when the stepper motor rotates, assuming that the current theoretical position of the stepper motor is , the target position is x, calculate the backlash error of the drive shaft at x ;like , keep the current position unchanged; if , it means that the stepper motor rotates in the forward direction, and the actual target position of the stepper motor is revised to ;like , it means that the stepper motor rotates in the opposite direction, and the actual target position of the stepper motor is revised to .

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] Compared with the traditional closed-loop control heliostat, the method of the present invention does not require an additional angle measurement device, thereby reducing the equipment cost. Compared with the traditional open-loop control heliostat, the backlash error can be compensated for each heliostat individually, thereby improving the transmission backlash compensation accuracy and solving the problem of backlash error changing during the life cycle of the heliostat. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of backlash error;

[0040] Figure 2 It is the flow chart of transmission backlash measurement and compensation;

[0041] Figure 3 This is a schematic diagram of the heliostat search area;

[0042] Figure 4 It is a schematic diagram of the position of the backlash error;

[0043] Figure 5 This is the principle diagram of backlash error measurement;

[0044] Figure 6 Schematic diagram of backlash error compensation. DETAILED DESCRIPTION

[0045] The following will be combined with the embodiments of the present invention and the drawings of the specification to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] The present invention proposes a method for measuring and compensating the transmission backlash error of a dual-axis heliostat using a light source and a camera. The method can continuously measure and compensate the transmission backlash error of each heliostat during the operation of the mirror field. Figure 2 As shown in the figure, it is mainly divided into four processes: preparation, error sampling, regression analysis and error compensation.

[0047] The preparation is to calibrate the camera and calculate the search area of ​​each heliostat in the camera.

[0048] The error sampling process is to measure the backlash error of the drive axis at multiple step positions, and each backlash error is called an error sample. The specific method is to select several different light source, heliostat and camera ternary combinations. Each ternary combination requires the heliostat to reflect the light source to the camera at different rotation angles. Each ternary combination can collect 2 error samples for each drive axis.

[0049] Regression analysis is the process of determining the relationship between the position of the drive shaft stepper motor and the backlash error. Different regression equations are selected according to the mechanical characteristics of the drive shaft and the factors affecting the error. Different drive shafts of the heliostat may require different regression equation forms. For example, the pitch drive shaft needs to consider the influence of gravity factors more than the horizontal drive shaft. After collecting enough error samples, the parameters of the equation can be fitted using regression analysis.

[0050] Error compensation refers to the process of adjusting the actual arrival position of the stepper motor based on the target position given by the control system, taking into account the backlash error at that position.

[0051] Compared with the traditional closed-loop control heliostat, the method of the present invention does not require an additional angle measurement device, thereby reducing the equipment cost. Compared with the traditional open-loop control heliostat, the backlash error can be compensated for each heliostat individually, thereby improving the transmission backlash compensation accuracy and solving the problem of backlash error changing during the life cycle of the heliostat.

[0052] A method for measuring and compensating a backlash error of a heliostat during operation of a mirror field comprises the following steps:

[0053] S0. Preparation.

[0054] 1) Calibrate the camera parameters. For specific methods, refer to "A Calibration Device for Tower-type Thermal Field Lens Camera", patent number "ZL 2023 2 1434711.0".

[0055] 2) Calculate the search area for the heliostat on the camera image. The search area is a rectangular pixel area that represents the possible location of the heliostat on the image. Figure 3 The search area calculation process is as follows: the heliostat is facing the camera, the camera imaging model is used to calculate the position of the heliostat vertex on the image, and then the minimum outer expansion rectangle of all vertex pixel positions is calculated as the theoretical search area, and finally a certain pixel range is expanded outward to obtain the final search area. The expansion takes into account the influence of factors such as camera lens jitter and camera calibration error, which lead to deviations in vertex pixel positions.

[0056] S1. Error sampling:

[0057] 1) A camera can collect backlash error samples for several heliostats at the same time. The specific number depends on the imaging parameters of the camera, the layout of the mirror field, etc. If the sun is used as the light source, the temperature resistance of the camera must also be considered. For the same camera, the search areas of the heliostats being sampled must not overlap to avoid incorrect identification of heliostats. The process of sampling any error is similar, and only the single error sampling process is described in detail below.

[0058] 2) Select a light source, heliostat, and camera combination, where the light source can be artificial light source or sunlight. The mirror normal vector when the heliostat reflects the light source to the camera is (x, y, z). To reduce the impact of the overfitting problem in regression analysis, the angle difference between the mirror normal of this combination and the mirror normal of any existing error sample is required to be no less than Angle Limit .

[0059] 3) Next, measure the backlash errors at the four positions in turn, see attached Figure 4 . First, the concept of reflection area is introduced. The reflection area is a step position range. For a given three-way combination of light source, heliostat and camera, when the step position is in the reflection area, it means that the light source can be reflected into the camera. It can be understood that the reflection area is the mapping of the reflection spot in the step position space. Taking the common dual-axis drive, rectangular heliostat as an example, the reflection area is a quadrilateral. The four boundary centers of the reflection area are selected as measurement positions, among which the upper and lower measurement positions V2 and V1 measure the backlash error of the pitch axis. The left and right measurement positions H1 and H2 measure the backlash error of the horizontal axis. Because the reflection area is a theoretical calculation value and the heliostat has tracking errors, the four selected positions are usually not strictly central positions, but this does not affect the backlash error measurement results.

[0060] 4) The measurement methods of the backlash errors at the four positions are similar. The following introduces the measurement process of the backlash error at a single position.

[0061] 5) Rotate the heliostat to the center of the reflection area. Ideally, the camera is at the center of the reflection spot.

[0062] 6) Fix the position of other drive shafts at a certain step interval δ e =0.1*E0 gradually reduces the current measured drive shaft step position, where E0 is the typical backlash error. Every time a position is moved, a photo is taken to identify the light spot and the light spot intensity is calculated. , spot intensity is the gray value sum of all spot pixels, as shown below, where n is the number of spot pixels, Gray k is the grayscale value of the kth spot pixel.

[0063] 。

[0064] 7) Continue to drive the stepper motor until the camera is at the edge of the reflected light spot. At this time, the light spot disappears. Record this position as , the distance the motor moves from the camera being at the center of the light spot to the camera being at the edge of the light spot is recorded as L1.

[0065] 8) Continue to drive the stepper motor for a distance L2=4*E0. L2 is to ensure that the camera moves out of the light spot range.

[0066] 9) Change the direction of the motor to e Move L1+L2 distance at intervals. Record the intensity of the light spot after each movement. .

[0067] 10) Calculate the backlash error. Figure 5 As shown, the solid circle curve represents the positive rotation position and the corresponding spot intensity The hollow circle curve represents the position of the reverse rotation and the corresponding spot intensity . Backlash error at The definition is as follows:

[0068] .

[0069] because is a curve represented by some discrete data. It is likely that there are no known discrete data points. In this case, linear interpolation is used to calculate a continuous position. Theoretical spot intensity at .

[0070] ;

[0071] in is less than The maximum known position of '' is greater than The smallest known position of .

[0072] 11) Repeat steps 5 to 10 to calculate the backlash errors at the remaining three locations in the reflection area.

[0073] 12) Evaluate the acquired backlash error. Because the two measurement positions of the same drive shaft in the reflection area are very close, the backlash error is measured. and It should also be very close. Calculate the relative error of two error samples of the same drive axis in the reflection area ,if If the error is greater than 0, the evaluation is passed, otherwise the two error samples need to be discarded. The calculation formula is as follows:

[0074] .

[0075] 13) Repeat steps 2 to 12 for a single heliostat until enough error samples are collected. The regression equation of backlash error and step position determines the minimum number of samples, denoted as N min .

[0076] S2. Regression analysis. The relationship between backlash error and step position is obtained. The regression analysis processes of different drive axes are independent of each other, so only the error analysis process of a single drive axis is introduced below.

[0077] 1) The backlash error data corresponding to several step positions are obtained through the error sampling process .

[0078] 2) Select the regression equation based on the mechanical structure of the drive and the factors affecting the backlash error, and convert the backlash error The relationship between the step position x is recorded as ,in is a set of parameter vectors for the regression equation.

[0079] 3) Use the least squares method to solve and find the parameter vector The best estimate of , so that the sum of squared errors of the samples is minimized:

[0080] .

[0081] 4) Calculate the coefficient of determination R 2 , evaluate the regression equation parameter vector Is it acceptable? 2 The calculation formula is as follows.

[0082] .

[0083] 5) If R 2 >R 2 limit , then accept the vector Otherwise, remove the sample with the largest deviation and repeat steps 1 to 4. If the number of samples is less than or equal to N min , it means that the existing samples cannot complete the regression analysis and it is necessary to continue collecting samples.

[0084] S4. Error compensation:

[0085] 1) Define three sets of stepper motor position and rotation angle relationship curves, see attached Figure 6 As shown. Among them, f2 is the relationship between the step position and the rotation angle when the stepper motor rotates in the forward direction. As the number of steps increases, the rotation angle increases. f1 is reverse rotation. As the number of steps decreases, the rotation angle decreases. Between f1 and f2 is the backlash error. f is located between f1 and f2, which is the theoretical relationship between the stepper motor and the rotation angle.

[0086] 2) When the stepper motor rotates, assuming that the current theoretical position of the stepper motor is , the target position is x, calculate the backlash error of the drive shaft at x .like , keep the current position unchanged; if , it means that the stepper motor rotates in the forward direction, and the actual target position of the stepper motor is revised to ;like , it means that the stepper motor rotates in the opposite direction, and the actual target position of the stepper motor is revised to .

[0087] 3) Drive the motor to move to the designed target position after error compensation.

[0088] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for measuring and compensating the backlash error of a heliostat during operation of a mirror field, characterized in that: The following steps are involved: S1. Calculate the heliostat search area on the camera image; S2. Error sampling, calculating backlash error, and evaluating the collected backlash error; For a given light source, heliostat, and camera combination, when the stepping position is in the reflection area, the light source is reflected into the camera, that is, the reflection area is the mapping of the reflected light spot in the stepping position space. The reflection area can measure four backlash errors. The single backlash error measurement process is as follows: S21. Rotate the heliostat to the center of the reflection area; S22. Fix the other drive axis positions at the set step interval Gradually reduce the current measured drive shaft step position, where E0 is the typical backlash error; take a photo to identify the light spot at each move, and calculate the light spot intensity , spot intensity is the sum of the grayscale values ​​of all light spot pixels, and the formula is as follows: ; Where n is the number of spot pixels, Gray k is the gray value of the kth spot pixel; S23. Continue to drive the stepper motor to make the camera at the edge of the reflected light spot. At this time, the light spot disappears. Record this position as , the distance the motor moves from the camera being at the center of the light spot to the camera being at the edge of the light spot is recorded as L1; S24. Continue to drive the stepper motor for a distance L2=4*E0; change the direction of the motor to δ e Move the distance L1+L2 at intervals and record the intensity of the light spot after each movement. ; Backlash error at The definition is as follows: ； The spot intensity is a curve represented by several discrete data, and a continuous position is calculated by linear interpolation. Theoretical spot intensity at: ； in is less than The maximum known position of '' is greater than The minimum known position of ; S25. Repeat the above steps to measure the backlash errors at other positions of the reflection area; S3. Regression analysis to obtain the relationship between backlash error and step position; S4. Perform error compensation and drive the motor to move to the designed target position after error compensation.

2. The method for measuring and compensating the backlash error of a heliostat during the operation of a mirror field according to claim 1, characterized in that: In step S1, the search area calculation process is as follows: the heliostat is facing the camera, the position of the heliostat vertex on the image is calculated using the camera imaging model, and then the minimum outer expansion rectangle of all vertex pixel positions is calculated as the theoretical search area, and finally the set pixel range is expanded outward to obtain the final search area.

3. The method for measuring and compensating the backlash error of a heliostat during the operation of a mirror field according to claim 1, characterized in that: Evaluate the acquired backlash error: Calculate the relative error of two error samples for the same drive axis in the reflection area ,if If the evaluation is passed, the two error samples need to be discarded. The calculation formula is as follows: ； Repeat steps S21 to S25 for a single heliostat until error samples that meet the requirements are collected. The regression equation of backlash error and step position determines the minimum number of samples, which is recorded as N. min .

4. The method for measuring and compensating the backlash error of a heliostat during the operation of a mirror field according to claim 1, characterized in that: In step S3: S31. The backlash error data corresponding to several step positions are obtained through the error sampling process. ; S32. Select the regression equation according to the mechanical structure of the drive and the factors affecting the backlash error, and convert the backlash error The relationship between the step position x is recorded as ,in is a set of parameter vectors of the regression equation; S33. Use the least squares method to solve and find the parameter vector The best estimate of , so that the sum of squared errors of the samples is minimized: ； S34. Calculate the coefficient of determination R 2 , evaluate the regression equation parameter vector Is it acceptable? 2 The calculation formula is as follows: ； S35. If R 2 >R 2 limit , then accept the vector Otherwise, remove the sample with the largest deviation and repeat steps S31 to S34. If the number of samples is less than or equal to N min , it means that the existing samples cannot complete the regression analysis and it is necessary to continue collecting samples.

5. The method for measuring and compensating the backlash error of a heliostat during the operation of a mirror field according to claim 1, characterized in that: In step S4, when the stepper motor rotates, it is assumed that the current theoretical position of the stepper motor is , the target position is x, calculate the backlash error of the drive shaft at x ;like , keep the current position unchanged; if , it means that the stepper motor rotates in the forward direction, and the actual target position of the stepper motor is revised to ;like , it means that the stepper motor rotates in the opposite direction, and the actual target position of the stepper motor is revised to .

Images