Optical coupling equipment alignment method based on machine vision

Abstract

The invention relates to an optical coupling equipment alignment method based on machine vision. The method comprises the following steps: extracting real-time parameter values of refractive index change, polarization fluctuation and phase delay according to optical data acquired by a system; inputting the extracted real-time parameter values of the refractive index change, the polarization fluctuation and the phase delay into a pre-established multi-dimensional parameter model for joint modeling; if the change of the refractive index exceeds a preset threshold, adjusting the focal length of the system, otherwise, keeping the current focal length; calculating an accumulated phase delay effect based on the recalculated parameter value, judging whether the phase delay affects the imaging definition or not, and if so, performing phase correction; re-evaluating the system performance index according to the optical data after phase correction so as to ensure the stability of the system; and optimizing the multi-dimensional parameter model based on the re-evaluated system performance indexes, and obtaining an optimal solution of the optical parameters through multiple iterative calculations.

Description

Technical Field

[0001] The present application relates to the technical field of laser measuring instruments, and in particular to a method for aligning optical coupling devices based on machine vision. Background Art

[0002] In machine vision and optical coupling systems, multi-dimensional aggregation optimization of optical parameters is a complex technical problem. During the alignment process, the system needs to simultaneously consider the dynamic changes of multiple optical parameters such as refractive index, polarization state, and phase delay. These parameters affect each other during zooming, making it difficult to achieve a balance between aberration and light energy loss.

[0003] Specifically, changes in the refractive index will affect the propagation path of light, while fluctuations in the polarization state will change the energy distribution of light, and differences in phase delay will lead to a decline in imaging quality. The joint modeling of these parameters requires finding the optimal solution in multi-dimensional space, but the coupling of the parameters makes the optimization process extremely complicated.

[0004] During zooming, as the focal length of the system increases, the gradient change in refractive index exacerbates aberrations, while anomalous fluctuations in the polarization state further amplify optical energy loss. At the same time, the cumulative effect of phase delay reduces image clarity. Furthermore, convergent control algorithms under multi-objective constraints must simultaneously meet multiple performance metrics. Achieving an optimal solution in this multi-dimensional parameter space remains a core challenge facing current technologies. Summary of the Invention

[0005] In order to solve the problems existing in the above-mentioned prior art, the purpose of this application is to provide an optical coupling device alignment method based on machine vision.

[0006] The present application discloses a method for aligning optical coupling devices based on machine vision, comprising the following steps:

[0007] Step S101, extracting real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on optical data collected by the system;

[0008] Step S102, inputting the extracted real-time parameter values ​​of refractive index variation, polarization fluctuation and phase delay into a multi-dimensional parameter model for joint modeling;

[0009] Step S103, calculating the mutual influence relationship between the refractive index change, polarization fluctuation and phase delay using a multi-dimensional parameter model, and determining whether the refractive index change exceeds a preset threshold based on the calculation result;

[0010] Step S104, if the refractive index change exceeds a preset threshold, adjust the system focal length, otherwise maintain the current focal length;

[0011] Step S105, recalculating the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay according to the optical data after the focal length adjustment;

[0012] Step S106, performing aberration compensation on the recalculated parameter values ​​using an aberration compensation algorithm to reduce light energy loss;

[0013] Step S107, calculating the cumulative phase delay effect based on the recalculated parameter value, determining whether the phase delay affects the imaging clarity, and performing phase correction if so;

[0014] Step S108 , re-evaluating system performance indicators based on the phase-corrected optical data to ensure system stability;

[0015] Step S109 , optimizing the multi-dimensional parameter model based on the re-evaluated system performance indicators, and obtaining the optimal solution of the optical parameters through multiple iterative calculations.

[0016] Preferably, in step S101, extracting real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on optical data collected by the system includes: acquiring real-time optical data collected by a group of optical sensors; performing feature extraction on the real-time optical data to obtain characteristic parameters characterizing the refractive index change, polarization fluctuation and phase delay; and outputting the characteristic parameters as real-time parameter values ​​to a multidimensional parameter model.

[0017] Preferably, in step S102, the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into a pre-established multidimensional parameter model for joint modeling, including: obtaining historical parameter values ​​of refractive index change, polarization fluctuation and phase delay; inputting the real-time parameter values ​​and the historical parameter values ​​into the multidimensional parameter model, and realizing joint modeling through the mapping relationship in the model to obtain a complete multidimensional optical parameter model.

[0018] Preferably, in step S103, the calculation of the mutual influence relationship between the refractive index change, polarization fluctuation and phase delay through the multidimensional parameter model includes: inputting the real-time parameter values ​​of the refractive index change, polarization fluctuation and phase delay into the multidimensional optical parameter model; calculating the functional relationship between the three based on the mathematical equation preset in the model; and determining the degree of mutual influence of the refractive index change, polarization fluctuation and phase delay according to the functional relationship.

[0019] Preferably, in step S105, aberration compensation is performed on the recalculated parameter values ​​using a preset aberration compensation algorithm, including: obtaining recalculated refractive index change, polarization fluctuation and phase delay parameter values; determining the degree of deviation between the main light and the ideal focal plane based on the above parameter values; selecting corresponding compensation coefficients in the aberration compensation algorithm for different degrees of deviation; and correcting the optical imaging data using the compensation coefficients to obtain compensated optical data.

[0020] Preferably, in step S107, the calculation of the cumulative phase delay effect based on the recalculated parameter value includes: obtaining a set of continuously collected refractive index changes, polarization fluctuations and phase delay parameter values; calculating the phase delay difference between two adjacent sets of parameter values; accumulating multiple sets of phase delay differences to obtain a cumulative phase delay amount; and judging whether the imaging clarity is affected based on the cumulative phase delay amount.

[0021] Preferably, in step S108, the re-evaluation of the system performance index based on the phase-corrected optical data includes: obtaining a set of test optical data after phase correction; extracting statistical characteristics of the test optical data, including mean, variance and signal-to-noise ratio; comparing the statistical characteristics with a preset performance index threshold; and judging whether the optical system meets the set stability requirements based on the comparison result.

[0022] Preferably, in step S109, the optimization of the multidimensional parameter model based on the re-evaluated system performance indicators includes: obtaining system performance indicator data evaluated multiple times; dynamically adjusting the weights of various parameters in the multidimensional parameter model according to the performance indicator data; searching for the optimal combination of model parameters through an iterative optimization algorithm; and generating a streamlined and efficient multidimensional optical parameter model using the optimal combination of model parameters.

[0023] The advantage of the machine vision-based optical coupling device alignment method described in the present application is that it extracts real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay by collecting optical data, and inputs them into a multidimensional parameter model for joint modeling. The present invention determines whether the refractive index change exceeds a preset threshold based on the model calculation results, and adjusts the system focal length accordingly. Subsequently, the present invention performs aberration compensation and phase correction on the adjusted optical data to reduce light energy loss and ensure imaging clarity. Finally, the present invention optimizes the multidimensional parameter model based on the re-evaluated system performance indicators, and obtains the optimal solution of the optical parameters through multiple iterative calculations. This method can realize real-time optimization of optical system parameters, effectively improve system performance and stability, and provide important support for high-precision optical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1This is the process of a method for aligning optical coupling devices based on machine vision described in this application Figure 1 ;

[0025] Figure 2 This is the process of a method for aligning optical coupling devices based on machine vision described in this application Figure 2 . DETAILED DESCRIPTION

[0026] like Figure 1-Figure 2 As shown, the optical coupling device alignment method based on machine vision described in this application includes the following steps:

[0027] Step S101, extracting real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on optical data collected by the system;

[0028] Step S102, inputting the extracted real-time parameter values ​​of refractive index variation, polarization fluctuation and phase delay into a multi-dimensional parameter model for joint modeling;

[0029] Step S103, calculating the mutual influence relationship between the refractive index change, polarization fluctuation and phase delay using a multi-dimensional parameter model, and determining whether the refractive index change exceeds a preset threshold based on the calculation result;

[0030] Step S104, if the refractive index change exceeds a preset threshold, adjust the system focal length, otherwise maintain the current focal length;

[0031] Step S105, recalculating the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay according to the optical data after the focal length adjustment;

[0032] Step S106, performing aberration compensation on the recalculated parameter values ​​using an aberration compensation algorithm to reduce light energy loss;

[0033] Step S107, calculating the cumulative phase delay effect based on the recalculated parameter value, determining whether the phase delay affects the imaging clarity, and performing phase correction if so;

[0034] Step S108 , re-evaluating system performance indicators based on the phase-corrected optical data to ensure system stability;

[0035] Step S109 , optimizing the multi-dimensional parameter model based on the re-evaluated system performance indicators, and obtaining the optimal solution of the optical parameters through multiple iterative calculations.

[0036] like Figure 1-Figure 2 As shown, in step S101, real-time parameter values ​​of refractive index variation, polarization fluctuation and phase delay are extracted based on optical data collected by the system.

[0037] Furthermore, in step S101, optical data is collected to obtain the original optical signal recorded by the system;

[0038] Analyze optical signals and separate basic information such as light intensity, wavelength and phase;

[0039] Calculate the initial refractive index change value based on the light intensity and wavelength data;

[0040] Extract polarization state information and analyze polarization fluctuation parameters of light waves;

[0041] Determine the phase delay value of the light wave through the phase information;

[0042] If the system focal length is adjusted, reacquire the adjusted optical data;

[0043] Recalculate the refractive index change value based on the adjusted optical data;

[0044] updating polarization fluctuation and phase delay parameters based on the adjusted optical data;

[0045] The obtained refractive index change, polarization fluctuation and phase delay parameters are input into the multidimensional parameter model for joint modeling.

[0046] Specifically, in step S101, optical data is collected to obtain the original optical signal recorded by the system, including collecting an optical signal with a wavelength of 532 nm through a spectrometer with a sampling frequency of 100 Hz;

[0047] Analyze optical signals, separate the basic information of light intensity, wavelength and phase, and use Fourier transform algorithm to decompose the signal into intensity spectrum and phase spectrum;

[0048] Based on the light intensity and wavelength data, the initial refractive index change value is calculated, and the refractive index is calculated using Snell's law in combination with the light intensity change, including a refractive index change of 0.0015 when the light intensity changes by 10%;

[0049] Extract polarization state information, analyze the polarization fluctuation parameters of the light wave, and calculate the polarization degree to be 0.85 and the polarization angle to be 45 degrees through the Stokes parameters;

[0050] The phase information is used to determine the phase delay of the light wave, and the phase unwrapping algorithm is used to obtain the phase delay as π / 2;

[0051] If the system focal length is adjusted, reacquire the adjusted optical data, including recollecting data after adjusting the focal length from 50mm to 75mm;

[0052] Based on the adjusted optical data, the refractive index change value was recalculated, and the refractive index change was obtained as 0.0018 using the corrected light intensity and wavelength data;

[0053] Based on the adjusted optical data, the polarization fluctuation and phase delay parameters are updated, and the polarization degree is recalculated to be 0.88 and the phase delay is π / 3;

[0054] The acquired refractive index change, polarization fluctuation and phase delay parameters are input into a multidimensional parameter model for joint modeling, and a multivariate linear regression algorithm is used to establish a relationship model between refractive index, polarization and phase.

[0055] like Figure 1-Figure 2 As shown, in step S102, the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into a multi-dimensional parameter model for joint modeling.

[0056] Furthermore, in step S102 , optical data is collected to extract original parameter values ​​of refractive index variation, polarization fluctuation, and phase delay;

[0057] Recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on the optical data after focus adjustment;

[0058] Perform data preprocessing on the extracted real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay to eliminate noise and outliers;

[0059] Inputting the pre-processed real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay into the input layer of the multi-dimensional parameter model;

[0060] The hidden layer of the multidimensional parameter model is used to extract features and perform joint modeling on the input parameter values ​​to obtain a multidimensional feature vector;

[0061] Based on the multidimensional feature vector, the weights and biases in the model are calculated to determine the correlation between the parameters;

[0062] Through the model output layer, the joint modeling results of refractive index change, polarization fluctuation and phase delay are obtained;

[0063] If the joint modeling results meet the preset threshold, a condition assessment report of the optical system is generated;

[0064] The parameter weights and biases in the multidimensional parameter model are updated according to the condition assessment report of the optical system.

[0065] Specifically, in step S102, when collecting optical data, a spectrometer is used to obtain light intensity distribution at a sampling frequency of 1000 times per second, and the original parameter values ​​of refractive index change, polarization fluctuation, and phase delay are extracted by Fourier transform, where the refractive index change range is 1.33 to 1.45, the polarization fluctuation amplitude is 0.01 to 0.05, and the phase delay range is 0 to 2π;

[0066] Based on the optical data after focal length adjustment, the Gaussian fitting algorithm is used to recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay to ensure that the data accuracy reaches 10 -4 ;

[0067] The extracted real-time parameter values ​​were preprocessed, noise was eliminated using wavelet transform, and outliers were removed using Z-score standardization method to make the data distribution mean 0 and variance 1;

[0068] The pre-processed real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay are input into the input layer of the multi-dimensional parameter model. The number of neurons in the input layer is 128, and the activation function is ReLU.

[0069] The hidden layer of the multidimensional parameter model is used to extract features and jointly model the input parameter values. The hidden layer contains three layers, with 256, 128, and 64 neurons in each layer, respectively. The back-propagation algorithm is used to optimize the weights to obtain a multidimensional feature vector.

[0070] Based on the multidimensional feature vector, the weights and biases in the model are calculated, and the gradient descent method is used to determine the correlation between the parameters. The learning rate is set to 0.001;

[0071] Through the model output layer, the Softmax function is used to obtain the joint modeling results of refractive index change, polarization fluctuation and phase delay. The number of neurons in the output layer is 3;

[0072] If the joint modeling results meet the preset thresholds, that is, the refractive index variation error is less than 0.001, the polarization fluctuation error is less than 0.005, and the phase delay error is less than 0.01, a status assessment report of the optical system is generated, which includes the parameter values ​​and their confidence intervals.

[0073] Based on the optical system's condition assessment report, the Adam optimizer is used to update the parameter weights and biases in the multidimensional parameter model, ensuring that the model accuracy is improved to 99.9.

[0074] like Figure 1-Figure 2 As shown, in step S103, the mutual influence relationship among the refractive index change, polarization fluctuation and phase delay is calculated by a multi-dimensional parameter model, and it is determined whether the refractive index change exceeds a preset threshold based on the calculation result.

[0075] Furthermore, in step S103, initial parameters of the optical system are obtained, including raw data of refractive index, polarization fluctuation and phase delay;

[0076] A multi-dimensional parameter model is used to model and calculate the mutual influence between refractive index change, polarization fluctuation and phase delay;

[0077] Extract the real-time value of refractive index change based on the calculation results of the multi-dimensional parameter model;

[0078] Compare the extracted refractive index change value with a preset threshold to determine whether it exceeds the threshold range;

[0079] If the refractive index change exceeds a preset threshold, the focal length parameters of the optical system are adjusted and the optical data is reacquired;

[0080] Recalculate the real-time values ​​of refractive index change, polarization fluctuation and phase delay according to the adjusted focal length parameters;

[0081] The recalculated refractive index change, polarization fluctuation and phase delay values ​​are input into the multi-dimensional parameter model for joint modeling;

[0082] Update the mutual influence relationship between refractive index change, polarization fluctuation and phase delay through joint modeling results;

[0083] According to the updated mutual influence relationship, the optimization parameters of the optical system are determined and the model calculation is completed.

[0084] Specifically, in step S103, the initial parameters of the optical system include a refractive index of 1.52, a polarization fluctuation of 0.03 rad, and a phase delay of 0.15 rad, and these raw data are acquired through a data acquisition module;

[0085] A multi-dimensional parameter model was used to input the initial values ​​of refractive index, polarization fluctuation, and phase delay. The interaction between the three was calculated using matrix operations and the least squares method, and the refractive index variation coefficient was obtained to be 0.02.

[0086] According to the calculation results of the multi-dimensional parameter model, the real-time value of the refractive index change is extracted as 0.021;

[0087] The extracted refractive index change value 0.021 is compared with the preset threshold value 0.02 to determine whether it exceeds the threshold range;

[0088] If the refractive index change exceeds a preset threshold, the algorithm adjusts the focal length parameters of the optical system from 10 mm to 10.5 mm and reacquires the optical data;

[0089] According to the adjusted focal length parameters, the real-time values ​​of refractive index change, polarization fluctuation and phase delay were recalculated using the finite element analysis method, which were 0.019, 0.028 rad and 0.14 rad respectively;

[0090] The recalculated values ​​are input into the multidimensional parameter model, and the Gaussian elimination method and iterative optimization algorithm are used for joint modeling;

[0091] By combining the modeling results, the mutual influence relationship between refractive index change, polarization fluctuation and phase delay is updated to obtain a new mutual influence coefficient matrix;

[0092] According to the updated mutual influence relationship, the gradient descent method is used to determine the optimization parameters of the optical system, complete the model calculation and output the final result.

[0093] like Figure 1-Figure 2 As shown, in step S104, if the refractive index change exceeds a preset threshold, the system focal length is adjusted, otherwise the current focal length is maintained.

[0094] Furthermore, in step S104, real-time parameter values ​​of refractive index variation, polarization fluctuation and phase delay of the current optical system are obtained through the multi-dimensional parameter model;

[0095] Based on the calculation results of the multi-dimensional parameter model, the mutual influence relationship between refractive index change, polarization fluctuation and phase delay is determined;

[0096] Comparing the preset threshold with the real-time parameter value of the refractive index change to determine whether the refractive index change exceeds the preset threshold;

[0097] If the refractive index change exceeds a preset threshold, the system focus adjustment mechanism is triggered and a focus adjustment instruction is generated;

[0098] If the refractive index change does not exceed the preset threshold, the current system focal length is kept unchanged, and the current state of the optical system is maintained;

[0099] According to the focus adjustment instruction, perform the system focus adjustment operation and recalibrate the focus parameters of the optical system;

[0100] Recalculate the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay after focal length adjustment through a multi-dimensional parameter model;

[0101] Use the recalculated real-time parameter values ​​to update the optical data in the multi-dimensional parameter model to ensure the real-time performance of the model parameters;

[0102] Based on the updated multidimensional parameter model, the dynamic changes of refractive index variation, polarization fluctuation and phase delay continue to be monitored.

[0103] Specifically, in step S104, the real-time parameter values ​​of the refractive index change, polarization fluctuation, and phase delay of the current optical system are obtained through the multidimensional parameter model, including using the finite element analysis method to calculate that the refractive index change is 1.52, the polarization fluctuation is 0.03, and the phase delay is 0.12;

[0104] Based on the calculation results of the multi-dimensional parameter model, the mutual influence relationship between the refractive index change, polarization fluctuation and phase delay was determined. The correlation coefficient between the refractive index change and polarization fluctuation was 0.85 using the linear regression model analysis.

[0105] A preset threshold is used to compare the real-time parameter value of the refractive index change to determine whether the refractive index change exceeds the preset threshold, including a preset threshold of 1.50. If the real-time refractive index change is 1.52, it exceeds the threshold;

[0106] If the refractive index change exceeds a preset threshold, the system focus adjustment mechanism is triggered, generating a focus adjustment instruction, including adjusting the focal length from 50mm to 52mm;

[0107] If the refractive index change does not exceed the preset threshold, the current system focal length is kept unchanged, and the current state of the optical system is maintained;

[0108] According to the focus adjustment instruction, perform the system focus adjustment operation and recalibrate the focus parameters of the optical system, including adjusting the focus to 52mm through the precision stepper motor;

[0109] The real-time parameter values ​​of the refractive index change, polarization fluctuation, and phase delay after focal length adjustment were recalculated using a multidimensional parameter model, including a corrected refractive index change of 1.50, a polarization fluctuation of 0.02, and a phase delay of 0.10;

[0110] Update the optical data in the multi-dimensional parameter model using the recalculated real-time parameter values ​​to ensure the real-time performance of the model parameters, including updating the refractive index change to 1.50;

[0111] Based on the updated multidimensional parameter model, the dynamic changes of refractive index changes, polarization fluctuations and phase delays continue to be monitored, including setting the monitoring frequency to 10 times per second to obtain optical parameter values ​​in real time.

[0112] like Figure 1-Figure 2 As shown, in step S105, the real-time parameter values ​​of the refractive index change, polarization fluctuation and phase delay are recalculated based on the optical data after the focus adjustment.

[0113] Furthermore, in step S105 , basic parameters of the optical system such as wavelength, incident angle, and beam intensity are extracted based on the optical data after the focal length adjustment;

[0114] Using optical transmission equations and combining basic parameters, calculate the propagation path and energy distribution of light beams in optical systems;

[0115] By using the refractive index calculation formula, combined with the beam propagation path and energy distribution, the real-time parameter value of the refractive index change is obtained;

[0116] According to the polarization state analysis method, the polarization state change information of the light beam during propagation is extracted; the polarization fluctuation model is used to calculate the real-time parameter value of the polarization fluctuation in combination with the polarization state change information;

[0117] The real-time parameter value of phase delay is obtained by combining the phase delay calculation formula with the beam propagation path and polarization state change information;

[0118] Integrate real-time parameter values ​​of refractive index variation, polarization fluctuation, and phase delay according to the input requirements of the multi-dimensional parameter model;

[0119] A joint modeling method of multidimensional parameter model is used to conduct multidimensional correlation analysis on the integrated real-time parameter values;

[0120] The comprehensive performance indicators of the optical system at different focal lengths are determined through the output results of the multidimensional parameter model.

[0121] Specifically, in step S105 , based on the optical data after the focal length is adjusted to 200 mm, basic parameters such as wavelength λ = 532 nm, incident angle θ = 30°, and beam intensity I = 100 mW are extracted;

[0122] Using the optical transmission equation and combining these parameters, the propagation path of the light beam in the optical system is calculated, and the energy distribution E(x,y)=I·exp(-(x 2 +y 2 ) / w 2 ), where w is the beam radius;

[0123] The refractive index calculation formula n = sinθ / sinθ' is used to combine the propagation path and energy distribution to obtain the real-time parameter value of the refractive index change Δn = 0.01;

[0124] According to the polarization state analysis method, the polarization state change information of the light beam during propagation is extracted, and the Stokes parameter S = [1, 0.8, 0.6, 0] is obtained;

[0125] The polarization fluctuation model is used in combination with the Stokes parameters to calculate the real-time parameter value of the polarization fluctuation ΔP = 0.02;

[0126] The phase delay calculation formula Δφ = (2π / λ)·Δn·d is used to combine the propagation path and polarization state change information to obtain a real-time parameter value of phase delay Δφ = 0.05 radians.

[0127] According to the input requirements of the multidimensional parameter model, the real-time parameter values ​​of the refractive index change Δn, polarization fluctuation ΔP and phase delay Δφ are integrated to form the input matrix M = [Δn, ΔP, Δφ];

[0128] The multidimensional correlation analysis of the input matrix was performed using the joint modeling method of the multidimensional parameter model, and the correlation coefficient matrix R = [[1, 0.9, 0.8], [0.9, 1, 0.85], [0.8, 0.85, 1]] was obtained;

[0129] The comprehensive performance index Q of the optical system at a focal length of 200 mm was determined to be 0.95 through the output results of the multi-dimensional parameter model.

[0130] like Figure 1-Figure 2 As shown, in step S106, an aberration compensation algorithm is used to perform aberration compensation on the recalculated parameter value to reduce light energy loss.

[0131] Furthermore, in step S106, real-time parameter values ​​of refractive index variation, polarization fluctuation, and phase delay are obtained based on the optical data after the focal length adjustment;

[0132] An aberration compensation algorithm is used to process the recalculated refractive index changes, polarization fluctuations, and phase delay parameter values;

[0133] Calculate the light energy loss distribution in the optical system through the aberration compensation algorithm;

[0134] According to the distribution of light energy loss, determine the compensation parameters that need to be adjusted in the aberration compensation algorithm;

[0135] Compensating the aberration in the optical system using compensation parameters to obtain compensated optical data;

[0136] Recalculate the cumulative phase delay effect based on the compensated optical data;

[0137] Determine whether the cumulative phase delay effect exceeds a preset threshold, and if so, perform phase correction;

[0138] Adjust the phase delay parameters in the optical system through the phase correction algorithm to obtain corrected optical data;

[0139] The imaging clarity information of the optical system is updated according to the corrected optical data.

[0140] Specifically, in step S106 , based on the optical data after the focal length adjustment, numerical simulation is performed to obtain a refractive index change of 1.45 to 1.55, a polarization fluctuation range of ±0.02, and a phase delay of 0.1 to 0.3 radians;

[0141] The Zernike polynomial aberration compensation algorithm is used to fit the recalculated refractive index change, polarization fluctuation and phase delay parameter values ​​to obtain the aberration distribution matrix.

[0142] Through the aberration compensation algorithm, the finite element analysis method is used to calculate the light energy loss distribution in the optical system, and the loss value is between 0.5% and 2.5%;

[0143] According to the distribution of light energy loss, the compensation parameters that need to be adjusted in the aberration compensation algorithm are determined, and the compensation coefficient is set to 0.8 to 1.2;

[0144] Compensation parameters are used to compensate for the aberrations in the optical system. The compensated optical data is obtained through iterative optimization, and the root mean square error of the aberration is reduced to below 0.05.

[0145] The accumulated phase delay effect was recalculated using Fourier transform based on the compensated optical data, and the phase delay value was 0.15 to 0.25 radians;

[0146] Determine whether the cumulative phase delay effect exceeds a preset threshold of 0.2 radians. If so, use the least squares method to perform phase correction;

[0147] The phase delay parameter in the optical system is adjusted by a phase correction algorithm, and the corrected phase delay value is controlled between 0.1 and 0.15 radians;

[0148] Based on the corrected optical data, the point spread function is used to update the imaging clarity information of the optical system, and the resolution is improved to 200 line pairs per millimeter.

[0149] like Figure 1-Figure 2 As shown, in step S107, the cumulative phase delay effect is calculated based on the recalculated parameter value to determine whether the phase delay affects the imaging clarity. If so, phase correction is performed.

[0150] Furthermore, in step S107, initial parameter values ​​of the optical system are obtained, including refractive index, polarization state and phase information;

[0151] Based on a multi-dimensional parameter model, the influence of refractive index changes on polarization fluctuations is calculated;

[0152] Determine the cumulative amount of phase delay effect based on the calculation results of polarization fluctuation;

[0153] The cumulative amount of phase delay effect is used to determine whether it affects the imaging clarity;

[0154] If the phase delay effect affects the imaging clarity, a phase correction algorithm is used to correct it;

[0155] Recalculate the refractive index change of the system based on the phase-corrected data;

[0156] Determine whether the recalculated refractive index change exceeds a preset threshold;

[0157] Re-evaluate system performance indicators based on the refractive index change not exceeding a preset threshold;

[0158] Determine the final stability of the system based on the re-evaluated system performance indicators.

[0159] Specifically, in step S107 , initial parameter values ​​of the optical system are obtained, including a refractive index of 1.45, a polarization state of 0.8, and a phase information of π / 3;

[0160] Based on a multi-dimensional parameter model, the finite element analysis method was used to calculate the influence of the refractive index change on the polarization fluctuation, and the amplitude of the polarization fluctuation was found to be 0.12;

[0161] According to the calculation results of polarization fluctuation, the cumulative amount of phase delay effect is determined to be 0.05π through the integration algorithm;

[0162] The clarity evaluation function is used to determine whether the phase delay effect affects the image clarity based on the cumulative amount of the effect, with a clarity threshold of 0.9.

[0163] If the phase delay effect affects the imaging clarity, the least square method is used for phase correction, and the corrected phase value is π / 4;

[0164] Based on the phase-corrected data, the refractive index change of the system was recalculated and the refractive index change value was 0.02;

[0165] The recalculated refractive index change is compared with a preset threshold of 0.03 to determine whether it exceeds the preset threshold;

[0166] Based on the refractive index change that did not exceed the preset threshold, the system performance indicators, including resolution and contrast, were re-evaluated, resulting in a resolution of 200 lp / mm and a contrast of 95%;

[0167] Based on the re-evaluated system performance indicators, the final stability of the system was determined to be 98%.

[0168] like Figure 1-Figure 2 As shown, in step S108, the system performance index is re-evaluated based on the phase-corrected optical data to ensure system stability.

[0169] Furthermore, in step S108, the phase-corrected optical data is acquired, and the phase-corrected optical signal is extracted;

[0170] Calculate the real-time parameter value of the system phase delay according to the phase-corrected optical signal;

[0171] Analyze the phase stability of the optical system using the real-time parameter value of the system phase delay;

[0172] Extracting real-time parameter values ​​of refractive index changes based on phase-corrected optical data;

[0173] The real-time parameter value of refractive index change is used to calculate the refractive index fluctuation range of the optical system;

[0174] According to the refractive index fluctuation range, the refractive index stability of the optical system under different environments is judged;

[0175] Using phase-corrected optical data, real-time parameter values ​​of polarization fluctuations are extracted;

[0176] Calculate the polarization stability index of the optical system based on the real-time parameter values ​​of polarization fluctuations;

[0177] The refractive index stability, phase stability and polarization stability indicators are used to comprehensively evaluate the overall performance of the optical system.

[0178] Specifically, in step S108, phase-corrected optical data is obtained from the optical system, and a Fourier transform algorithm is used to extract the phase-corrected optical signal, including by calculating a phase component of a signal with a frequency of 500 Hz;

[0179] Based on the extracted phase correction signal, the least square method is used to calculate the real-time parameter value of the system phase delay, including the delay time of 0.05ms;

[0180] Based on the phase retardation parameter value, the phase stability of the optical system is analyzed by calculating the phase standard deviation of 0.01 rad;

[0181] From the phase-corrected optical data, the refractive index calculation formula is used to extract the real-time parameter value of the refractive index change, including the refractive index change range of 1.45 to 1.47;

[0182] Using the refractive index variation parameter value, the refractive index fluctuation range of the optical system is obtained by calculating the refractive index fluctuation amplitude as 0.02;

[0183] Based on the refractive index fluctuation range, the refractive index stability of the optical system under different environments is judged, including the refractive index fluctuation is less than 0.01 when the temperature changes by ±5℃;

[0184] From the phase-corrected optical data, the real-time parameter values ​​of polarization fluctuations were extracted through Stokes parameters, including the degree of polarization of 0.95;

[0185] Based on the polarization fluctuation parameter value, the polarization stability index of the optical system is calculated, including the polarization angle deviation less than 0.5°;

[0186] Combining the refractive index stability, phase stability and polarization stability indicators, a weighted comprehensive evaluation method is used to calculate the overall performance score of the optical system, including a comprehensive score of 95 points.

[0187] like Figure 1-Figure 2 As shown, in step S109, the multi-dimensional parameter model is optimized based on the re-evaluated system performance indicators, and the optimal solution of the optical parameters is obtained through multiple iterative calculations.

[0188] Furthermore, in step S109, the phase-corrected optical data is acquired and the system performance index is extracted;

[0189] Based on the extracted system performance indicators, an initial parameter set of the multi-dimensional parameter model is constructed;

[0190] An iterative calculation method is used to adjust the optical parameters in the multi-dimensional parameter model;

[0191] The interaction between refractive index change, polarization fluctuation and phase delay is calculated through a multi-dimensional parameter model;

[0192] Determining whether the calculated refractive index change exceeds a preset threshold;

[0193] If the refractive index change does not exceed a preset threshold, the optical parameters in the multidimensional parameter model are updated according to the calculation results;

[0194] Repeat the iterative calculation process to perform multiple iterative optimizations on the updated multi-dimensional parameter model;

[0195] Through multiple iterative calculations, the optimal solution of optical parameters is obtained;

[0196] Re-evaluate system performance indicators based on the optimal solution and confirm system stability.

[0197] Specifically, in step S109 , system performance indicators are extracted from the phase-corrected optical data, including extracting a root mean square (RMS) value of the wavefront error of 0.05λ and a Strehl ratio of 0.92;

[0198] Based on the extracted system performance indicators, the initial parameter set of the multidimensional parameter model was constructed, including the initial value of the refractive index of 1.45, the polarization fluctuation range of ±0.1°, and the initial value of the phase delay of π / 4;

[0199] An iterative calculation method, including the gradient descent method, was used to adjust the optical parameters in the multidimensional parameter model, with a learning rate of 0.01 and an iteration step of 100 times.

[0200] The interplay between refractive index variation, polarization fluctuation, and phase retardation is calculated using a multidimensional parameter model, including the use of matrix operations to solve the covariance matrix of refractive index variation and polarization fluctuation.

[0201] Determine whether the calculated refractive index change exceeds a preset threshold, where the preset threshold is ±0.01;

[0202] If the refractive index change does not exceed the preset threshold, the optical parameters in the multidimensional parameter model are updated according to the calculation results, including adjusting the refractive index to 1.452 and correcting the polarization fluctuation range to ±0.08°;

[0203] Repeat the iterative calculation process to perform multiple iterations of optimization on the updated multi-dimensional parameter model, and record the parameter convergence status at each iteration;

[0204] Through multiple iterative calculations, the optimal solution for the optical parameters was obtained, including a final refractive index of 1.453, a polarization fluctuation range of ±0.07°, and a phase delay of π / 3;

[0205] The system performance indicators were re-evaluated based on the optimal solution, including the calculated wavefront error root mean square value reduced to 0.03λ and the Strehl ratio increased to 0.95, confirming the system stability.

[0206] Those skilled in the art can make various other corresponding changes and deformations based on the technical solutions and concepts described above, and all of these changes and deformations should fall within the scope of protection of the claims of this application.

Claims

1. A method for aligning optical coupling devices based on machine vision, characterized in that: The following steps are involved: Step S101, extracting real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay based on optical data collected by the system; Step S102, inputting the real-time parameter values ​​of refractive index variation, polarization fluctuation, and phase delay extracted in step S101 into a multi-dimensional parameter model for joint modeling; Step S103, calculating the mutual influence relationship between the refractive index change, polarization fluctuation and phase delay using the multidimensional parameter model in step S102, and determining whether the refractive index change exceeds a preset threshold based on the calculation result; Step S104, if the refractive index change exceeds a preset threshold, adjust the system focal length, otherwise maintain the current focal length; Step S105, recalculating the real-time parameter values ​​of refractive index change, polarization fluctuation and phase delay according to the optical data after the focal length adjustment in step S104; Step S106, performing aberration compensation using an aberration compensation algorithm for the parameter values ​​recalculated in step S105 to reduce light energy loss; Step S107, calculating the cumulative phase delay effect based on the parameter value recalculated in step S106, determining whether the phase delay affects the imaging clarity, and performing phase correction if so; Step S108, re-evaluating the system performance index based on the optical data after phase correction in step S107; Step S109 , optimizing the multi-dimensional parameter model based on the system performance indicators re-evaluated in step S108 , and obtaining the optimal solution of the optical parameters through iterative calculation.

2. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: Extracting the real-time parameter value in step S101 includes: acquiring real-time optical data collected by the optical sensor, separating the light intensity, wavelength and phase information through Fourier transform, calculating the refractive index change value in combination with Snell's law, analyzing the polarization fluctuation using Stokes parameters, and determining the phase delay value through a phase unwrapping algorithm.

3. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: The joint modeling in step S102 specifically includes: The real-time parameter values ​​and historical parameter values ​​are input into the multidimensional parameter model, and the functional relationship between the refractive index, polarization and phase is established through the multivariate linear regression algorithm, and the model weights are optimized based on the gradient descent method.

4. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: Adjusting the system focal length in step S103 includes: When the refractive index change exceeds a threshold, a focus adjustment instruction is generated and the focus is adjusted by a stepper motor within a range of ±5% of the current focus value. After adjustment, optical data is recollected for parameter verification.

5. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: The aberration compensation algorithm in step S104 is a Zernike polynomial algorithm, including: The aberration distribution matrix is ​​fitted according to the refractive index change, polarization fluctuation and phase delay parameter values, the light energy loss distribution is calculated by finite element analysis, and the compensated optical data is iteratively optimized based on the compensation coefficient.

6. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: The phase correction in step S105 includes: The cumulative phase delay effect was corrected by the least squares method, the phase delay value was controlled within ±10% of the preset threshold, and the imaging clarity was updated by the point spread function.

7. The optical coupling device alignment method based on machine vision according to claim 1, characterized in that: Optimizing the multidimensional parameter model in step S106 includes: Based on the system performance indicators after phase correction, the Adam optimizer is used to dynamically adjust the model parameter weights, and the gradient descent method is used for iterative optimization until the parameter convergence error is less than 0.5%.

8. The optical coupling device alignment method based on machine vision according to claim 2, characterized in that: The sampling frequency of the Fourier transform is 100 Hz to 1000 Hz, and the calculation accuracy of the Stokes parameter is ±0.

01.

9. The optical coupling device alignment method based on machine vision according to claim 3, characterized in that: The hidden layer of the multidimensional parameter model contains three layers of neural networks, with the number of neurons being 256, 128, and 64 respectively, the activation function being ReLU, and the output layer using the Softmax function to generate the joint modeling result.

10. The optical coupling device alignment method based on machine vision according to claim 5, characterized in that: The compensation coefficient of the Zernike polynomial algorithm ranges from 0.8 to 1.2, and the root mean square error of the aberration after compensation is less than 0.05.

Images