CN115933159A - Real-time high-precision wavefront distortion phase compensation system - Google Patents

Real-time high-precision wavefront distortion phase compensation system Download PDF

Info

Publication number
CN115933159A
CN115933159A CN202211349096.3A CN202211349096A CN115933159A CN 115933159 A CN115933159 A CN 115933159A CN 202211349096 A CN202211349096 A CN 202211349096A CN 115933159 A CN115933159 A CN 115933159A
Authority
CN
China
Prior art keywords
wavefront
distortion phase
real
deformable mirror
processing module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211349096.3A
Other languages
Chinese (zh)
Inventor
刘宇韬
郑明伟
徐苗
付广伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yanshan University
Original Assignee
Yanshan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yanshan University filed Critical Yanshan University
Priority to CN202211349096.3A priority Critical patent/CN115933159A/en
Publication of CN115933159A publication Critical patent/CN115933159A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

The invention discloses a real-time high-precision wavefront distortion phase compensation system which comprises a deformable mirror, a beam splitter, a first focusing lens, a second focusing lens, a receiving module, a CCD (charge coupled device) camera, a U-Net convolution neural network processing module and a wavefront reconstruction module; the light intensity information of the light beam with the distorted phase is collected through the CCD camera, the U-Net network processing module carries out real-time high-precision sensing on the wavefront distorted phase, and the wavefront reconstruction module generates corresponding driving voltage according to the wavefront distorted phase information to enable the deformable mirror to deform so as to compensate the wavefront distorted phase.

Description

Real-time high-precision wavefront distortion phase compensation system
Technical Field
The invention relates to a real-time high-precision wavefront distortion phase compensation system, and belongs to the field of wavefront-free sensing adaptive optics.
Background
With the rapid development of science and technology, the conventional radio frequency communication system cannot meet the rapid demand of the social development for high-speed communication services in terms of communication rate or data capacity. Free space coherent optical communication is considered as an important technical means for breaking through the bottleneck of the existing high-speed communication due to the outstanding technical advantages of high communication speed and large information capacity. However, in practical applications, the optical signal may suffer from negative effects such as light intensity flicker, light beam drift, etc. caused by atmospheric turbulence, which seriously affect the communication quality of the free space coherent optical communication system.
Compared with the conventional adaptive optics, the wavefront-free sensing adaptive optics has the greatest characteristic that a wavefront sensor is not used for detecting wavefront distortion, but a wavefront corrector is directly controlled by an optimization algorithm to optimize signal light, and the method becomes one of the main research directions for inhibiting atmospheric turbulence and improving the communication quality of a free-space coherent optical communication system in recent years. S.W.Paine and J.R.Fienup et al, both at Rochester university, abroad, use a wavefront-free sensing adaptive optics system based on a traditional IncepotionV 3 network architecture to effectively predict wavefront aberrations. Y.nishizaki et al, osaka university, japan, accurately calculated the first 32 Zernike coefficients using CNN-based neural networks. In China Caoshitai and the like, a random parallel Gradient Descent (SPGD) algorithm is used for correcting wavefront aberration in a free space coherent optical communication system, and simulation and experiment results show that the SPGD algorithm can effectively compensate wavefront distortion and improve the communication quality of the free space coherent optical communication system. Horse comet sensitivity and the like provide a wavefront-free sensing algorithm based on a convolutional neural network, and simulation results show that the Steckel ratio of light spots is obviously improved after distortion compensation. The above method has the following problems:
1. the atmospheric turbulence model responsible for providing the training sample is mostly simulated by using a method based on Zernike polynomials, the turbulence sample generated by the method has the problem of insufficient high-frequency information, and the network model trained by using the samples is easy to have low sensing precision, so that the wavefront distortion phase cannot be accurately compensated.
2. Most of the optimization algorithms responsible for controlling the wavefront reconstruction module use traditional convolutional neural networks, and the network models have the problem of weak data processing capacity and cannot realize real-time compensation of wavefront distortion phases.
Therefore, it is a technical problem to be solved urgently to design a system for carrying out real-time accurate correction of wavefront distortion phase by adopting a wavefront-free sensing algorithm.
Disclosure of Invention
The invention aims to solve the technical problems and provides a real-time high-precision wavefront distortion phase compensation system which is simple in structure, low in cost, strong in data processing capacity of a distortion phase correction algorithm, high in sample precision used by network training and beneficial to realizing real-time high-precision compensation of wavefront distortion phases.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a real-time high-precision wavefront distortion phase compensation system comprises a deformable mirror, a beam splitter, a first focusing lens, a second focusing lens, a receiving module, a CCD (charge coupled device) camera, a U-Net convolution neural network processing module and a wavefront reconstruction module;
the wave front reconstruction module generates corresponding driving voltage according to wave front distortion phase information extracted by the U-Net network processing module, and the deformable mirror generates deformation under the control of the driving voltage to correct distortion wave front and further outputs a modulation optical signal; the modulated optical signals corrected by the deformable mirror pass through the beam splitter, one path of the modulated optical signals passes through the first focusing lens to reach the receiving module, and the other path of the modulated optical signals passes through the second focusing lens to be collected by the CCD camera; light intensity image data acquired by the CCD camera is input into the U-Net network processing module; the U-Net network processing module carries out iterative updating on network parameters by using a sample error loss function formed by the predicted distortion phase information and the actual distortion phase information, and stops iterative updating until a set maximum iteration number K is reached;
the sample error loss function of the U-Net network processing module for iterative updating is as follows:
Figure BDA0003918257220000021
Figure BDA0003918257220000022
Figure BDA0003918257220000031
wherein the evaluation index is j, relu (x) = max (0, x); p is a radical of (j) Representing the actual distortion phase;
Figure BDA0003918257220000032
representing a predicted distortion phase;
predicting distorted phase
Figure BDA0003918257220000033
The specific expression of (A) is as follows:
Figure BDA0003918257220000034
the smaller the error is, the closer the predicted distortion phase is to the actual distortion phase, and finally the average of all sample errors in a training set is used for measuring the prediction quality of the network model, wherein the specific expression is as follows:
Figure BDA0003918257220000035
wherein m is the number of samples.
The technical scheme of the invention is further improved as follows: setting the initial driving voltage vector output by the wave front reconstruction module as
Figure BDA0003918257220000036
The drive voltage vector is iteratively updated using equation (6):
v (k+1) ＝v (k) +γΔv (k) ΔJ (k) (6)，
wherein v is (k+1) 、v (k) Respectively obtaining driving voltage vectors of the k +1 th iteration and the k th iteration; gamma is a positive gain coefficient;
Figure BDA0003918257220000037
for the disturbance voltage vector, Δ v, generated at the kth iteration (k) Each element in the Chinese character is subjected to Bernoulli distribution, the absolute value of each element is fixed, and the probability of taking the positive sign and the negative sign is 1/2; delta J (k) The variation of the performance index J is calculated according to the formula (7):
Figure BDA0003918257220000038
Figure BDA0003918257220000039
the performance index changes in the positive direction and the negative direction are respectively calculated according to formulas (8) and (9):
Figure BDA0003918257220000041
Figure BDA0003918257220000042
wherein J [ v ] (k) +Δv (k) ]And J [ v ] (k) -Δv (k) ]Respectively is a performance index target function when the voltage changes towards the positive direction and the negative direction; j [ v ] (k) ]Is a performance index objective function of the driving voltage vector of the kth iteration; and the deformable mirror deforms according to the drive voltage after iteration, so that the distorted phase of the wavefront is compensated.
The technical scheme of the invention is further improved as follows: the deformable mirror adopts a 19-unit deformable mirror, a 21-unit deformable mirror, a 32-unit deformable mirror or a 45-unit deformable mirror.
The technical scheme of the invention is further improved as follows: the number m of samples is more than or equal to 70000.
The technical scheme of the invention is further improved as follows: the network parameters are convolution kernel parameters and scalar deviations of the network.
The technical scheme of the invention is further improved as follows: the maximum iteration number K is more than or equal to 4000.
The technical scheme of the invention is further improved as follows: the U-Net network processing module uses massive accurate samples to train the accurate samples in advance and optimizes the accurate samples by adopting a residual block, wherein the massive accurate samples are from an accurate atmospheric turbulence simulation model based on power spectrum inversion method simulation.
The technical scheme of the invention is further improved as follows: the above-mentioned
Figure BDA0003918257220000043
Is 1V-1.5V.
The technical scheme of the invention is further improved as follows: the gain coefficient gamma is 1.2-1.6.
The technical scheme of the invention is further improved as follows: the described
Figure BDA0003918257220000044
Is 0.2V-0.3V.
Due to the adoption of the technical scheme, the invention has the technical progress that:
the invention accurately compensates the wavefront distortion phase in real time through a wavefront-free sensing closed-loop correction system based on a U-Net network; the CCD camera is responsible for collecting light intensity information (usually, light spot brightness value or light spot energy distribution graph), which is easier to obtain than wavefront aberration; the U-Net network and the wavefront reconstruction module generate corresponding driving voltage according to the collected light intensity information to control the deformation of the deformable mirror and generate corresponding compensation phases; in the process of the distorted phase compensation, a wavefront sensor is not needed, but a U-Net network trained by massive accurate samples is used for inverting the phase distortion of the light beam through the light intensity information of the light beam and realizing the distorted phase compensation, and the method has the advantages of small volume, low cost and simple structure; meanwhile, the U-Net network has strong real-time data processing capacity, so that the real-time high-precision compensation of the wavefront distortion phase can be realized.
Drawings
FIG. 1 is a general system block diagram of the present invention;
FIG. 2 is a 32 unit anamorphic mirror driver layout of the present invention;
FIG. 3 is a schematic diagram of a U-Net network incorporating a residual block according to the present invention;
the device comprises a deformable mirror 1, a deformable mirror 2, a beam splitter 3, a first focusing lens 4, a second focusing lens 5, a receiving module 6, a CCD camera 7, a U-Net convolution neural network processing module 8 and a wavefront reconstruction module.
Detailed Description
In order to make the technical solution, advantages and objects of the present invention more clear and definite, the present invention will be described in detail with reference to the accompanying drawings and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a detailed operation process are given, but the scope of the present invention is not limited to the following embodiments.
With reference to fig. 1 to 3, the real-time high-precision wavefront distortion phase compensation system of the invention comprises a deformable mirror 1, a beam splitter 2, focusing lenses 3 and 4, a receiving module 5, a CCD camera 6, a u-Net convolution neural network processing module 7 and a wavefront reconstruction module 8, wherein the CCD camera 6 is a CCD scientific research-grade camera. The wavefront reconstruction module 8 generates corresponding driving voltage according to the wavefront distortion phase information extracted by the U-Net network processing module 7, and the deformable mirror 1 generates deformation correction distortion wavefront under the control of the driving voltage so as to output a modulation optical signal; one path of the modulated optical signal corrected by the deformable mirror passes through the first focusing lens 3 by the beam splitter 2 to reach the receiving module 5, and the other path of the modulated optical signal passes through the second focusing lens 4 and is collected by the CCD camera 6; the light intensity image data collected by the CCD camera 6 is input into a U-Net network processing module 7; and the U-Net network processing module 7 performs iterative updating on the network parameters by using a sample error loss function formed by the predicted distortion phase information and the actual distortion phase information, and stops iterative updating until the set maximum iteration number K is reached, wherein K is more than or equal to 4000.
The deformable mirror 1 may be a deformable mirror of 19 units, 21 units, 32 units, 45 units, or the like. This embodiment uses the 32-element deformable mirror shown in fig. 2.
The U-Net convolution neural network shown in FIG. 3 inputs wavefront distortion phase information collected by a scientific research CCD camera, and the input distortion phase information is subjected to a series of down-sampling and convolution encoding and a series of up-sampling and deconvolution decoding processes, and then turbulence distortion phase information extracted by the network is finally output. The number of network layers can be increased by adding the U-Net network of the residual block, and the middle copying jump connection layer is used for improving the extraction capability of network characteristics. The method adopts massive accurate samples provided by the atmospheric turbulence model simulated based on the power spectrum inversion method to train the network, and improves the accuracy of distortion phase compensation.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

1. The utility model provides a real-time high accuracy wavefront distortion phase compensation system which characterized in that: the device comprises a deformable mirror (1), a beam splitter (2), a first focusing lens (3), a second focusing lens (4), a receiving module (5), a CCD camera (6), a U-Net convolution neural network processing module (7) and a wavefront reconstruction module (8);
the wave front reconstruction module (8) generates corresponding driving voltage according to wave front distortion phase information extracted by the U-Net network processing module (7), and the deformable mirror (1) generates deformation correction distortion wave front under the control of the driving voltage so as to output a modulation optical signal; the modulated light signals corrected by the deformable mirror (1) pass through the beam splitter (2), one path of the modulated light signals reaches a receiving module (5) through the first focusing lens (3), and the other path of the modulated light signals is collected by the CCD camera (6) after passing through the second focusing lens (4); light intensity image data acquired by the CCD camera (6) is input into the U-Net network processing module (7); the U-Net network processing module (7) carries out iterative update on network parameters by using a sample error loss function formed by the predicted distortion phase information and the actual distortion phase information, and stops iterative update until a set maximum iteration number K is reached;
the sample error loss function of the iterative update of the U-Net network processing module (7) is as follows:
Figure FDA0003918257210000011
Figure FDA0003918257210000012
Figure FDA0003918257210000013
wherein the evaluation index is j, relu (x) = max (0, x); p is a radical of formula (j) Representing the actual distortion phase;
Figure FDA0003918257210000014
representing a predicted distortion phase;
predicting distorted phase
Figure FDA0003918257210000015
The specific expression of (A) is as follows:
Figure FDA0003918257210000016
the smaller the error is, the closer the predicted distortion phase is to the actual distortion phase, and finally the average of all sample errors in a training set is used for measuring the prediction quality of the network model, wherein the specific expression is as follows:
Figure FDA0003918257210000021
wherein m is the number of samples.
2. The real-time high-precision wavefront distortion phase compensation system of claim 1The method is characterized in that: setting the initial driving voltage vector output by the wave front reconstruction module (8) as
Figure FDA0003918257210000022
The drive voltage vector is iteratively updated using equation (6):
v (k+1) ＝v (k) +γΔv (k) ΔJ (k) (6)，
wherein v is (k+1) 、v (k) Respectively obtaining driving voltage vectors of the k +1 th iteration and the k th iteration; gamma is a positive gain coefficient;
Figure FDA0003918257210000023
for the disturbance voltage vector, Δ v, generated at the kth iteration (k) Each element in the sequence is subjected to Bernoulli distribution, the absolute value of each element is fixed, and the probability of taking the sign is 1/2; delta J (k) The variation of the performance index J is calculated according to the formula (7):
Figure FDA0003918257210000024
Figure FDA0003918257210000025
the performance index changes in the positive direction and the negative direction are respectively calculated according to formulas (8) and (9) to obtain:
Figure FDA0003918257210000026
Figure FDA0003918257210000027
wherein J [ v ] (k) +Δv (k) ]And J [ v ] (k) -Δv (k) ]When the voltage changes in positive and negative directions, respectivelyA performance index objective function; j [ v ] (k) ]Is a performance index objective function of the driving voltage vector of the kth iteration; the deformable mirror (1) deforms according to the drive voltage after iteration, so that the distorted phase of the wavefront is compensated.
3. The system of claim 1, wherein the system comprises: the deformable mirror (1) adopts a 19-unit, 21-unit, 32-unit or 45-unit deformable mirror.
4. The system of claim 1, wherein the system comprises: the number m of samples is more than or equal to 70000.
5. The system of claim 1, wherein the system comprises: the network parameters are convolution kernel parameters and scalar deviations of the network.
6. The system of claim 1, wherein the system comprises: the maximum iteration number K is more than or equal to 4000.
7. The system of claim 1, wherein the system comprises: the U-Net network processing module uses massive accurate samples to train the accurate samples in advance and optimizes the accurate samples by adopting a residual block, wherein the massive accurate samples are from an accurate atmospheric turbulence simulation model based on power spectrum inversion method simulation.
8. A real-time high precision wavefront distortion phase compensation system as claimed in claim 2 wherein: the described
Figure FDA0003918257210000031
Is 1V-1.5V.
9. A real-time high precision wavefront distortion phase compensation system as claimed in claim 2 wherein: the gain coefficient gamma is 1.2-1.6.
10. A real-time high precision wavefront distortion phase compensation system as claimed in claim 2 wherein: the above-mentioned
Figure FDA0003918257210000032
Is 0.2V-0.3V. />
CN202211349096.3A 2022-10-31 2022-10-31 Real-time high-precision wavefront distortion phase compensation system Pending CN115933159A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211349096.3A CN115933159A (en) 2022-10-31 2022-10-31 Real-time high-precision wavefront distortion phase compensation system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211349096.3A CN115933159A (en) 2022-10-31 2022-10-31 Real-time high-precision wavefront distortion phase compensation system

Publications (1)

Publication Number Publication Date
CN115933159A true CN115933159A (en) 2023-04-07

Family

ID=86699662

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211349096.3A Pending CN115933159A (en) 2022-10-31 2022-10-31 Real-time high-precision wavefront distortion phase compensation system

Country Status (1)

Country Link
CN (1) CN115933159A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117639452A (en) * 2024-01-23 2024-03-01 深圳市科沃电气技术有限公司 Voltage compensation method, device and equipment of inverter and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117639452A (en) * 2024-01-23 2024-03-01 深圳市科沃电气技术有限公司 Voltage compensation method, device and equipment of inverter and storage medium
CN117639452B (en) * 2024-01-23 2024-04-23 深圳市科沃电气技术有限公司 Voltage compensation method, device and equipment of inverter and storage medium

Similar Documents

Publication Publication Date Title
CN109031654B (en) Adaptive optical correction method and system based on convolutional neural network
CN110207835A (en) A kind of wave front correction method based on out-of-focus image training
CN1987547B (en) Device for automatic correcting telescope astigmatic aberration using telescope second lens
CN108983412A (en) A kind of no Wave-front measurement adaptive optics system and beam phase method of adjustment
CN103901617A (en) Wavefront detection-free adaptive optical system based on model
CN115933159A (en) Real-time high-precision wavefront distortion phase compensation system
CN109739033A (en) A kind of liquid-crystal aberration correcting method of no Wavefront detecting
CN111579097B (en) High-precision optical scattering compensation method based on neural network
CN106526839A (en) Mode-based synchronous wavefront sensorless adaptive optical system
CN114004342A (en) Laser communication system distortion wavefront prediction method based on LSTM network
CN109407311A (en) A kind of diffraction optics imaging system wavefront inversion algorithm based on phase difference method
CN116992779A (en) Simulation method and system of photovoltaic energy storage system based on digital twin model
CN112099229B (en) High-speed self-adaptive optical closed-loop control method based on far field
Archinuk et al. Mitigating the nonlinearities in a pyramid wavefront sensor
CN207882564U (en) Based on real-time phase difference technology without Wavefront detecting adaptive optics system
CN114994859B (en) TMA telescope surface shape and pose error on-orbit correction method based on CNN
CN115753018A (en) Advanced prediction correction method for distortion vortex light beam under dynamic turbulence
CN115471428A (en) Co-phase error correction method based on all-optical diffraction neural network
CN115598980A (en) Self-adaptive optical wavefront prediction and feedforward correction method based on deep learning
CN116400495B (en) Wavefront correction system based on RUN optimization algorithm
CN110068973B (en) Liquid crystal aberration correction method based on deconvolution neural network
Li et al. Experimental Study on Wavefront Distortion Correction of 40-Elements Adaptive Optical System Based on SPGD Algorithm
CN115857157B (en) Wave front-free sensing self-adaptive optical correction method of SPGD algorithm based on AMSGrad
CN109870800A (en) A kind of no Wavefront detecting adaptive optics system and its quick non-model control method
Liu et al. Residual network-based aberration correction in a sensor-less adaptive optics system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination