Interference signal correction and calculation method based on high-scribed-line density grating

Abstract

The invention relates to an interference signal correction and calculation method based on a high-scribed-line-density grating. The method comprises the steps that 1, four paths of grating interference signal data are collected through a photoelectric detector; 2, subtracting a self mean value from four paths of grating interference original signals received by each photoelectric detector, and performing direct current elimination processing; 3, carrying out filtering processing on the grating interference signal after the direct current elimination processing by adopting a mean value smooth filtering method; 4, setting a temporary memory for the filtered four-path grating interference signals for stack type storage, and setting a data acquisition and preprocessing module and an interference signal correction module to be in dual-thread synchronous operation; 5, correcting the displacement nonlinearity of the grating interference signal read from the temporary memory by using a correction algorithm based on RANSAC-Heydemann; and 6, calculating the displacement. According to the method, a double-closed-loop thread processing mode is designed and is innovatively combined with an RANSAC-Heydemann correction algorithm, so that the measurement stability and accuracy are greatly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for processing grating interference signals in the field of optical precision measurement using grating measurement method for length measurement, in particular discloses an interference signal correction and calculation method based on high line density grating. BACKGROUND

[0002] High line density grating interferometer is a high precision length measurement equipment directly taking grating period as reference, its core technology lies in taking high line density grating as key components and length reference, using Doppler effect generated by high line density grating vector direction movement to make two coherent lights on grating produce frequency difference, form beat frequency signal captured by photoelectric detector. Compared with traditional laser interferometer, high line density grating interferometer has significant technical advantages: the reference material grating changes little under the influence of temperature, has excellent thermal stability, and can maintain high precision measurement performance in complex environment. At present, the line density of high line density grating can reach 4500l / mm, which has more fine measurement resolution, and provides reliable technical foundation for micro-nano scale precision measurement. This kind of interferometer has important application value in precision manufacturing, semiconductor processing, optical detection, micro-electro-mechanical system and other fields. The period of high line density grating is much smaller than the wavelength of traditional laser interferometer, so that it can obtain more interference fringes under the same measurement stroke, thereby improving the measurement resolution. However, due to the special physical properties of high line density grating, its interference signal faces many challenges in actual measurement, such as signal quality affected by environmental factors, significant nonlinear error and other problems, which seriously affect the accuracy and stability of measurement. Therefore, it is a key problem to be solved in this technical field to effectively correct the interference signal of high line density grating and improve the signal quality and measurement accuracy.

[0003] The traditional single-frequency laser interferometer signal processing method mainly includes DC elimination, filter preprocessing, and signal correction schemes such as difference, orthogonalization and Heydemann correction. However, these methods have obvious shortcomings when processing high-line-density grating interference signals: first, the signal-to-noise ratio characteristics of high-line-density grating interference signals have a significant impact on measurement accuracy. In actual applications, due to the high grating line density and small period, the signal is easily disturbed by external interference, and the traditional least squares fitting method is prone to overfitting when processing such signals, resulting in distorted measurement results. Second, the non-ideal performance of the optical and optoelectronic devices used by the interferometer and their adjustment methods further exacerbate the non-linear error of the measurement results. High-line-density gratings are more sensitive to environmental changes, and factors such as temperature fluctuations and mechanical vibrations can introduce additional measurement errors. Traditional gross error discrimination methods such as the Leavitt criterion and the Romanovsky criterion are based on statistical principles and are suitable for error discrimination of direct measurement values, but they have obvious limitations when dealing with complex measurement tasks involving the functional relationship between multiple quantities, such as interference displacement calculation. In addition, existing signal processing methods also have shortcomings in terms of real-time performance and computational efficiency. At high sampling frequencies, the processing of large amounts of data can cause program iteration delays, affecting the real-time performance and stability of the measurement, making it difficult to meet the real-time measurement requirements of industrial sites. SUMMARY

[0004] To solve the problems existing in the prior art, the present application provides an interference signal correction and calculation method based on high-line-density gratings. The method first acquires four grating interference signals using photoelectric detectors; then performs DC elimination on the four grating interference signals by subtracting the mean value of each photoelectric detector from the original signal received by the photoelectric detector; then performs filter processing on the grating interference signals using a mean value smoothing filter; then performs stack storage on the four grating interference signals, sets up a double closed-loop program with a temporary register, and runs the data acquisition and preprocessing and interference signal correction modules in double threads; then performs nonlinear correction on the four grating interference signals using a RANSAC-Heydemann-based correction algorithm, including steps such as downsampling, differencing to two channels, establishing a Heydemann correction model, calculating the noise-to-signal ratio, setting the total set and random samples, sampling and estimating instances, calculating geometric distances, and selecting the largest consistent set; and finally calculating the displacement.

[0005] The present application is implemented as follows: an interference signal correction and calculation method based on high-line-density gratings, comprising the following steps: Step 1: Raw data acquisition, four grating interference signals are acquired using photoelectric detectors; Step 2: DC removal, the original signal received by each photodetector and collected by the collection card is subtracted by its own mean value to remove the DC component of the four-channel grating interference signal collected in step 1; Step 3: smoothing filtering, the four-channel grating interference signal obtained in step 2 is smoothed and filtered; Step 4: pre-processing data into stack, the filtered four-channel grating interference signal obtained in step 3 is stored in a stack, a temporary storage is set to compile a double closed loop processing procedure, data collection and preprocessing and interference signal correction are double-threaded and synchronized, each thread is used for data collection and preprocessing, and the data is continuously stored in the temporary storage, and the other thread is used for interference signal correction module to continuously read data from the temporary storage; Step 5: nonlinear correction, the interference signal correction module corrects the four-channel grating interference signal data read from the temporary storage for nonlinear correction, and the RANSAC-Heydemann correction algorithm is used to correct the nonlinear displacement of the grating interference signal; Step 6: displacement calculation, the corrected two-channel interference signal is calculated for displacement.

[0006] The specific steps of the step 5 based on the RANSAC-Heydemann correction algorithm for correcting the grating interference signal are as follows: Step 5.1: down-sampling the read four-channel grating interference signal, down-sampling the interference signal to filter out repeated samples and save computing power; Step 5.2: differentiating the four-channel grating interference signal after down-sampling in step 5.1 to two channels, and using the differential amplification method to remove the cooperative random error in real-time measurement process; Step 5.3: establishing a Heydemann correction model for the two-channel grating interference signal after differentiation in step 5.2; Step 5.4: calculating the ratio of static noise to working signal of the two-channel grating interferometer according to the Heydemann correction model in step 5.3; Step 5.5: setting the total set D of grating interference signals and the random sample J of the set data; Step 5.6: sampling and estimating the instance F from the total set D p (J) and the consistent set S[F p (J)]; Step 5.7: calculating the geometric distance of each element of the sample J to the function F p (J); Step 5.8: repeating steps 5.6 and 5.7 multiple times until the consistent set S[F p (J) with the maximum number of data points is selected, the model F is re-estimated with S[F p (J), and finally the result is output.

[0007] The displacement calculation specific steps in the step 6 are as follows: Step 6.1: arctangent, the two-way signal obtained after step 5 correction is directly divided, and the instantaneous phase of the division result is calculated by arctangent; Step 6.2: unwrapping, the phase splicing is realized by eliminating the step of instantaneous phase interval Pi; Step 6.3: phase-displacement conversion, further, the displacement of grating vector direction is calculated by combining the proportion coefficient of grating period d as the reference.

[0008] The photoelectric detector in the step 1 includes a photodiode and an amplification circuit connected with each other, the collected light signal is converted into an electrical signal through the photodiode, and then is amplified into a voltage analog signal through an analog circuit.

[0009] The smoothing filter in the step 3 is a mean value smoothing filter method for filtering the grating interference signal, a mean value sampling frame is set first, and then the average value of the neighborhood sampling points is used to replace the value of the pixel point in the mean value sampling frame.

[0010] The application provides reliable technical support for the practical application of the high-line-density grating interferometer by the innovative combination of the double closed-loop thread processing mode and the RANSAC-Heydemann correction algorithm, and promotes the application of the high-line-density grating technology in the precision measurement field.

[0011] Compared with the prior art, the application has the following beneficial effects: 1. The double closed-loop thread processing mode is adopted in the application, the double threads are synchronously operated, the CPU resources are fully utilized, the program iteration delay is reduced, the measurement real-time performance and the processing efficiency are significantly improved, and the real-time measurement requirement of the industrial field is met.

[0012] 2. The RANSAC-Heydemann correction algorithm is adopted in the application to intelligently remove gross errors, avoid overfitting, reduce the signal standard deviation by at least 50%, and greatly improve the measurement stability and accuracy, so that a reliable technical scheme is provided for precision measurement. DETAILED DESCRIPTION

[0013] Figure 1 It is a step flow schematic diagram of the interference signal correction and calculation method based on the high-line-density grating.

[0014] Figure 2 It is a smoothing filter processing schematic diagram in the interference signal correction and calculation method based on the high-line-density grating.

[0015] Figure 3This is a schematic diagram of dual-thread real-time measurement in the interference signal correction and solution method based on high line density grating described in this invention.

[0016] Figure 4 (a) to 4(d) are the consensus sets S[F] in the RANSAC-Heydemann modified algorithm of this invention. p A schematic diagram of the acquisition and processing process and output results of (J)].

[0017] Figure 5 This is a schematic diagram of the correction results of an embodiment of the interference signal correction and solution method based on a high line density grating described in this invention. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0019] According to the appendix Figure 1 This invention relates to a method for correcting and resolving interference signals based on high-line-density gratings, comprising the following steps: Step 1: Raw data acquisition. Four-channel grating interference signals are acquired using a photodetector. The photodetector consists of photodiodes and an amplifier circuit that are interconnected. The photodiodes convert the optical signals into electrical signals, which are then amplified into analog voltage signals by the analog circuit.

[0020] Step 2: DC component removal. DC cancellation is performed on the four grating interference signals acquired in Step 1. The mean value of each signal received by the photodetector and acquired by the acquisition card is subtracted from the original signal.

[0021] Step 3: Smoothing Filtering. The four-channel grating interference signals obtained in Step 2 after DC cancellation are smoothed and filtered. The mean smoothing filter method is used to filter the grating interference signals.

[0022] Step 4: Preprocessed data is pushed onto the stack. The filtered four-channel grating interference signals obtained in Step 3 are stored in a stacked manner. This invention sets up a dual-closed-loop processing flow with a temporary register, running the data acquisition and preprocessing and interference signal correction modules synchronously in two threads. On one hand, the acquired and preprocessed data is continuously and cyclically stored into the temporary register; on the other hand, the interference signal correction module continuously reads data from the temporary register. Each module uses one thread, making full use of the CPU's computing space, effectively reducing program iteration delays, and increasing the real-time performance of the grating interferometer measurement.

[0023] Step 5: Nonlinear correction. Read the four grating interference signals stored in the temporary register in Step 4 and perform nonlinear correction on them. Use the RANSAC-Heydemann correction algorithm to correct the displacement nonlinearity of the grating interference signals.

[0024] Step 5.1: downsampling the read four-channel grating interferometer signal. The interference signal is downsampled to filter out repeated samples and save computing power.

[0025] Step 5.2: differentiating the four-channel grating interferometer signal downsampled in step 5.1 into two channels. The method of differential amplification is used to remove the cooperative random error in real-time measurement.

[0026] Step 5.3: establishing a Heydemann correction model for the two-channel grating interferometer signal differentiated in step 5.2.

[0027] Step 5.4: calculating the ratio of the static noise of the two-channel grating interferometer to the working signal according to the Heydemann correction model in step 5.3.

[0028] Step 5.5: setting the total set D of grating interferometer signals and the random sample J of the set data.

[0029] Step 5.6: sampling and estimating the instance F in the total set D p (J) and the consistent set S[F p (J)].

[0030] Step 5.7: calculating the geometric distance of each element of the sample J to the function F p (J).

[0031] Step 5.8: repeating steps 5.6 and 5.7 multiple times until the consistent set S[F p (J)] with the maximum number of data points is selected, and the model F is re-estimated with S[F p (J)], and finally the result is output.

[0032] Step 6: displacement calculation, displacement calculation is performed on the corrected two-channel interference signal.

[0033] Step 6.1: arctangent, directly dividing the corrected two-channel interference signal, and then arctangent calculating the unwrapped instantaneous phase.

[0034] Step 6.2: unwrapping, eliminating the phase jump of the instantaneous phase interval Π to realize phase splicing.

[0035] Step 6.3: phase-displacement conversion, further, combined with the period d of the grating as the reference scale factor, the displacement of the grating vector direction can be calculated.

[0036] The pre-processing, signal correction and displacement solving algorithm software of the grating interferometer signal collected by the photoelectric detector adopts LabVIEW software of National Instruments Corporation, and a double-closed-loop thread processing mode is designed. Figure 3, the double closed loop thread processing mode, a thread is data extraction and preprocessing, using the stack mode maximum sampling rate of photodetector received grating interference signal acquisition, after AC-DC estimation preliminary elimination interference generated by the direct current component, and the data into the stack temporary storage. Two threads for interference signal correction module, for self tracing grating interferometer signal processing and calculation, the thread consists of signal correction and displacement solution two modules. Signal calibration module through adjustable sampling selection appropriate sample total set D, then through the difference, random increment algorithm to weaken random noise, large noise, with Heydemann correction further refined correction signal, eliminate nonlinear error; displacement solution module includes inverse tangent solution instantaneous phase, unwrapping phase splicing and through the phase-displacement conversion and accumulation finally calculate the self tracing grating interferometer measurement displacement value.

[0037] Embodiment: The specific working steps of the interference signal correction and solution method based on high line density grating according to the present application are as follows: Step 1: data acquisition, four grating interference signals are collected by photodetectors, the photodetectors are composed of photodiodes and amplification circuits, the photodiodes convert optical signals into electrical signals, and the analog circuits amplify the electrical signals into voltage analog signals. In this embodiment, the MCC USB 1608GX-2AO data acquisition card of National Instruments Corporation is used to convert the analog signals collected by the photodetectors into digital signals, and the maximum sampling frequency is 500ks / s, which can realize asynchronous sampling of 8 ports at the same time.

[0038] Step 2: direct current elimination of the collected four grating interference signals. Subtract the mean value of each photodetector received and collected by the acquisition card from the original signal. For the direct current voltage elimination of the grating interference signal, this example performs direct current preliminary screening. The principle of direct current preliminary screening is to subtract the mean value of each photodetector received and collected by the acquisition card from the original signal. Let the voltage value collected by the acquisition card of a photodetector be PD i , i = 1,2,3,4, then n original interference signals PD_o i,n After direct current preliminary screening, it is denoted as PD_dc i,n : (1).

[0039] Step 3: smooth filtering of the four grating interference signals after direct current preliminary screening, the mean smooth filtering method is used to filter the grating interference signals. The high frequency noise h1(t) generated by high speed jitter of grating and environmental disturbance is generally eliminated by filtering. In this example, the mean smooth filtering method is used to filter the signal after direct current preliminary screening, and the smooth filtering function in LabVIEW has high matrix operation capability.

[0040] Mean smoothing filtering first defines a mean sampling frame d, and within this frame, the average value of the neighboring sampling points is used to replace the pixel value. (See attached...) Figure 2 If the smoothing region is set to d=3 (i.e., the mean sampling frame, which is the thick black box area shown in the figure), and the step size is 1, the algorithm will slide and calculate the mean on the first row of the original signal. The smoothed data will then be... p Through formula (2) (2), Obtain the second row of data after smoothing and filtering.

[0041] Step 4: The four-channel grating interferometer signals after smoothing and filtering are stored in a stack. This invention employs a dual-loop program compiled with a temporary register, running the data extraction and preprocessing, and interferometer signal correction modules synchronously in two threads. On one hand, the results of data extraction and preprocessing are continuously and cyclically stored in the temporary register; on the other hand, the interferometer signal correction module continuously reads data from the temporary register. Each module uses its own thread, fully utilizing the CPU's computing space, effectively reducing program iteration delays, and increasing the real-time performance of the grating interferometer measurement.

[0042] Step 5: Correct the nonlinearity of the four-channel grating interference signals read in Step 4 by using the RANSAC-Heydemann correction algorithm to correct the displacement nonlinearity of the grating interference signals. The specific steps are as follows: Step 5.1: Downsampling. This invention sets the acquisition card to full sampling mode to maximize the acquisition of instantaneous signal characteristics. However, to filter out duplicate samples and save computing power, downsampling of the signal is necessary. Let the array before downsampling be... The downsampled array is The downsampling algorithm can be expressed as: (3), Where: n is the number of elements in X, m is the downsampling factor, s is the starting index, and l is the number of elements in the output sequence downsampling array. It's a floor operation, which takes the largest integer less than or equal to that number.

[0043] Step 5.2: The four grating interference signals obtained after downsampling in Step 5.1 are differentially divided into two channels, and differential amplification is used to remove the cooperative random error in the real-time measurement process. The four interference light intensity signals received by the photodetector are respectively phased by 3λ / 4, λ / 4, λ, and λ / 2 from the initial phase. Let the DC bias of each interference sine wave be b, the amplification factor be a, and the phase be θ. Its mathematical expression can be defined as: (4).

[0044] After downsampling in step 5.1, the four grating interference signals pd1 and pd2, and pd3 and pd4 are differentiated. Then, at time t, the expressions for the two sets of signals can be defined as follows: (5).

[0045] Step 5.3: Perform RANSAC-Heydemann correction on the two grating interference signals after differential processing in Step 5.2.

[0046] A Heydemann correction model is established for the differentiald two-path grating interference signals. In equation (5), and This determines the coordinates of the center of the ellipse; and The amplitudes of the differential signals together determine the major and minor axes of the ellipse; This represents the initial phase at time t; and This represents the random error of each of the two sets of differential signals; This represents the phase change when the grating vector direction is displaced. According to equation (5), we can obtain... The expression is: (6).

[0047] Clearly, to solve for the phase, we need to determine the five parameters in equation (6). , , , and According to the classic Heydemann correction model, the mutually orthogonal u(t) and v(t) signals are combined in space as the x-axis and y-axis, respectively, to form a spiral. The projection of this spiral onto the xOy axis is a circle or ellipse. Based on this, the constrained Heydemann correction mathematical model can be defined as: (7), In equation (7), F(u,v;X) is the Heydemann modified theoretical model, X is the model parameter to be fitted, u and v are the interference differential signals input on the x-axis and y-axis respectively, and D is the set of signal data points collected.

[0048] Step 5.4: Calculate the ratio of static noise to working signal of the two grating interferometers based on the Heydemann correction model described in Step 5.3. To avoid overfitting using the least squares method and increase the matching degree between the fitting result and the acquired samples, this invention combines the RANSAC algorithm to eliminate gross errors. First, calculate the ratio of static noise to working signal of the two grating interferometers.

[0049] Let the voltage values ​​collected by the i-th photodetector during the operation of the grating interferometer be as follows: i=1,2,3,4, when not in operation If i = 1, 2, 3, 4, then the ratio η of the static noise to the working signal of the grating interferometer can be expressed as: (8), S i SN represents the power of the n signals acquired by the i-th PD during the operation of the grating interferometer. i This represents the ambient noise power inside the channel when the grating interferometer is not operating.

[0050] Step 5.5: Set the total set D of the grating interference signals and the random sample J of the set data.

[0051] Let D be the set of differential signal data points obtained from the differential grating interferometer. Based on the noise signal ratio, let k = (1-η) be the number of randomly sampled elements required to solve the objective function F. n. A subset of k random data points is called a random sample J of the objective function F.

[0052] Step 5.6: Sample and estimate instance F from the total set D. p (J) and the uniform set S[F p (J)]. A sample J is randomly drawn from the total set D, and an estimation function F is obtained from this sample using the least squares method. p (J), determined with F p The set of data points whose geometric distance between (J) is less than the threshold ζ is called the estimated function F. p The uniform set of (J) is denoted as S[F]. p (J)].

[0053] Step 5.7: Calculate the function F for each element of sample J. p (J) geometric distance.

[0054] The distance function represents the distance from each element of sample J to the function F. p The geometric distance of (J) is given by the threshold ζ, which is the distance from the elements of sample J to the estimated function F. p (J) Maximum distance. Given that the distance from the focus of an ellipse to the ellipse is a constant, find the maximum distance using the elements of sample J to F. p The sum of the distances between the foci F1 and F2 of (J) and the constant value (F) p The difference between (J) and its major axis length (2a) is used as a threshold function, i.e.: (9), In the formula, line1 represents the elements from sample J to F. p(J) is the distance from the focus F1, and line2 is the distance from the element of sample J to F. p The distance of (J) to the focus F2.

[0055] Step 5.8: As shown in Figures 4(a) to 4(c), repeat steps 5.6 and 5.7 multiple times until the consistent set S[F] with the largest number of data points is selected. p (J)], in the figure, the thin-lined hollow circles represent the collected signal elements in the total set D, and the thick-lined hollow circles represent the uniform set S[F]. p (J)] element, solid black dots are the estimated function F p (J) element, finally use S[F p (J)] Re-estimate model F, output as follows Figure 4 (d) Results shown by the solid coil.

[0056] Step 6: Displacement calculation. Perform displacement calculation on the corrected two interference signals. The specific steps are as follows: Step 6.1: Arctangent. Divide the two signals directly, then use the arctangent of the division result to calculate the ununtangled instantaneous phase. This invention further arctangents the Heydemann-corrected signal to obtain the phase difference Δt: (10), In the formula S a S is the differential signal between pd1 and pd2 after correction in step 5. b This is the differential signal between pd3 and pd4 after step 5 correction.

[0057] Step 6.2: Unwinding. Unwinding is equivalent to phase stitching. The default domain for the computer's inverse tangent operation is [0, 2π]. This results in a step interval of π between the calculated displacement values. The unwrap algorithm eliminates these π-interval steps to achieve phase stitching. The unwrap algorithm expands the phase array by removing discontinuous parts. If the difference between adjacent values ​​in the phase is greater than π, and the phase unit is radians for input and output, the unwrap algorithm calculates the expanded phase using the following equation: (11), p out (i) represents the output phase result, where p is the phase and n is the phase length. It's a floor operation, rounding down.

[0058] Step 6.3: Phase-Displacement Conversion. Further, combining the period d = 212.882 nm of the self-traceable grating, the displacement in the grating vector direction can be obtained, i.e., the displacement Δx measurement expression of the self-traceable grating interferometer: (12) The interference signal from the grating interferometer was acquired using a Thorlabs silicon ribbon amplified photodetector (PDA10A2) and an MCC acquisition card (MCC 1608GX-2AO). The former can detect laser wavelengths in the range of 200nm-1000nm, while the latter can achieve an acquisition speed of 500Kstamp / s. The RANSAC-based Heydemann correction algorithm described in this invention was compiled using MATLAB software, with the distance function set as the distance from the elements of sample J to F. p The sum of the distances between the foci F1 and F2 of (J) and the constant value (F) p The difference between (J) and its major axis length twice that of 2a), with the threshold ζ being the element and F. p (J) The distance from the critical value of the top 50% is obtained Figure 5 The results are shown.

[0059] from Figure 5 In the RANSAC-based Heydemann correction algorithm, only points closer to the fitted curve are used as fitted sample points (red points), while points farther away from the fitted curve (blue points) are removed. The circle formed by the element set obtained by using the RANSAC-based Heydemann correction is more stable. To obtain specific comparison results, the mean and variance of the minimum distance between the fitted curve and the elements of the sample set were calculated for both RANSAC-processed and non-RANSAC-processed cases, as shown in equations (7) and (8). Five sets of data were collected, and the results are as follows: (13), (14), In the formula, Mean is the average of the minimum distances between the curve and the elements of the sample set, and x D,i It is the i-th sample element, x F,i is the point on the curve that is the minimum distance to the sample element, n is the total number of samples, and Var is the variance of the minimum distance between the curve and the elements of the sample set.

[0060] The comparison between RANSAC-processed and unprocessed samples is shown in Table 1 below:

[0061] As can be seen from Table 1 above, after using the interference signal correction and solution method based on high line density grating described in this invention, the average distance from the total sample set D to the fitted curve is significantly reduced after using the external points with a long distance. The standard deviation of the signal is reduced by at least 50%, that is, the fitted curve is more accurate, which greatly improves the stability of the signal.

[0062] Based on the above preparations, with the maximum motion range of the displacement stage set to 5000nm and the frequency of triangular wave motion to 1Hz, the interference signal correction and solution method based on a high line density grating described in this invention was used to process the grating interference signal. Continuous motion of the displacement stage was recorded for 5 seconds, and the results were compared with those not processed using the method described in this invention, as shown in Table 2 below:

[0063] It is evident that the method of this invention can intelligently eliminate gross errors, avoid overfitting, and significantly improve measurement stability and accuracy.

Claims

1. A method for correcting and solving interference signals based on high line density gratings, characterized in that... The interference signal correction and solution method includes the following steps: Step (1): Raw data acquisition, using a photodetector to acquire four-channel grating interference signals; Step (2): DC component removal. DC cancellation is performed on the four grating interference signals acquired in step (1). The original signal received by each photodetector and acquired by the acquisition card is reduced by its own mean. Step (3): Smoothing filtering, smoothing filtering the four-way grating interference signal obtained in step (2) after DC cancellation; Step (4): Preprocessed data is pushed onto the stack. The filtered four-channel grating interference signals obtained in step (3) are stored in a stack. A dual closed-loop processing flow with a temporary register is set up. The two modules of data acquisition and preprocessing and interference signal correction are run synchronously in two threads. Each module uses one thread. One thread continuously and cyclically stores the data after data acquisition and preprocessing into the temporary register, while the other thread continuously reads data from the temporary register for the interference signal correction module. Step (5): Nonlinear correction. The interference signal correction module performs nonlinear correction on the four-channel grating interference signal data read from the temporary storage, and corrects the displacement nonlinearity of the grating interference signal using the RANSAC-Heydemann correction algorithm. Step (6): Displacement calculation, perform displacement calculation on the corrected two interference signals.

2. The method for correcting and solving interference signals based on high line density gratings according to claim 1, characterized in that: The specific steps of step (5), which involves correcting the grating interference signal based on the RANSAC-Heydemann correction algorithm, are as follows: Step (5.1): Downsample the read four-channel grating interference signal to filter out duplicate samples and save computing power; Step (5.2): The four grating interference signals downsampled in step (5.1) are differentially divided into two channels, and differential amplification is used to remove the cooperative random error in the real-time measurement process; Step (5.3): Establish a Heydemann correction model for the two grating interference signals after differential processing in step (5.2); Step (5.4): Calculate the ratio of static noise to working signal of the two-channel grating interferometer according to the Heydemann correction model described in step (5.3); Step (5.5): Set the total set D of the grating interference signals and the random sample J of the set data; Step (5.6): Sample and estimate instances F from the total set D. p (J) and the uniform set S[F p (J)]; Step (5.7): Calculate the function F for each element of sample J. p (J) geometric distance; Step (5.8): Repeat steps (5.6) and (5.7) multiple times until the consistent set S[F] with the largest number of data points is selected. p (J)], using S[F p (J)] Re-estimate model F, and finally output the results.

3. The method for correcting and solving interference signals based on high line density gratings according to claim 1, characterized in that: The specific steps for displacement calculation in step (6) are as follows: Step (6.1): Arctangent, after correcting and processing the two signals obtained in step (5), perform a direct division operation, and then calculate the instantaneous phase of the unwound signal by arctangenting the division result; Step (6.2): ​​Unwrap the phases to eliminate the step transitions of the instantaneous phases in the interval Π and achieve phase splicing; Step (6.3): Phase-displacement conversion. Further, the displacement of the grating vector direction is calculated by combining the scaling factor based on the grating period d.

4. The method for correcting and solving interference signals based on high line density gratings according to claim 1, characterized in that: The photodetector in step (1) includes a photodiode and an amplifier circuit that are interconnected. The photodiode converts the collected light signal into an electrical signal, which is then amplified into a voltage analog signal by the analog circuit.

5. The method for interferometric signal correction and calculation based on a high line density grating according to claim 1, characterized in that: The smoothing filtering in step (3) is to use the mean smoothing filtering method to filter the grating interference signal. First, a mean sampling frame is set, and within the mean sampling frame, the average value of the neighboring sampling points is used to replace the value of the pixel.

Images