自适应光学系统几种随机并行优化控制算法比较

 直接对系统性能指标进行优化是自适应光学系统中一种重要的波前畸变校正方法,选择合适的随机并行优化控制算法是该技术成功实现的关键。以32单元变形镜为校正器,基于多种随机并行优化算法建立自适应光学系统仿真模型。从算法的收敛速度、校正效果、局部极值3个方面对遗传算法、单向扰动随机并行梯度下降、双向扰动随机并行梯度下降及模拟退火算法进行了比较。仿真结果表明,遗传算法收敛速度太慢,不适用于需要实时控制的自适应光学系统;双向扰动随机并行梯度下降算法收敛速度、校正效果要优于单向扰动随机并行梯度下降,且能够适应各种情况下的扰动电压;模拟退火几乎以概率1收敛到全局极值附近,且收敛速度是上述算法中最快的。     

Comparison of several stochastic parallel optimization algorithms for adaptive optics system without a wavefront sensor

Optimizing the system performance metric directly is an important method for correcting wavefront aberrations in an adaptive optics (AO) system where wavefront sensing methods are unavailable or ineffective. An appropriate “Deformable Mirror” control algorithm is the key to successful wavefront correction. Based on several stochastic parallel optimization control algorithms, an adaptive optics system with a 61-element Deformable Mirror (DM) is simulated. Genetic Algorithm (GA), Stochastic Parallel Gradient Descent (SPGD), Simulated Annealing (SA) and Algorithm Of Pattern Extraction (Alopex) are compared in convergence speed and correction capability. The results show that all these algorithms have the ability to correct for atmospheric turbulence. Compared with least squares fitting, they almost obtain the best correction achievable for the 61-element DM. SA is the fastest and GA is the slowest in these algorithms. The number of perturbation by GA is almost 20 times larger than that of SA, 15 times larger than SPGD and 9 times larger than Alopex.     

基于Zernike模式的自适应光学系统随机并行梯度下降算法

 控制算法的收敛速度一定程度上限制了无波前探测自适应光学技术在实时波前畸变校正中的应用。从理论分析角度提出将模式法和区域法结合起来以提高算法收敛速度,并以61单元变形镜为校正器,建立基于随机并行梯度下降算法自适应光学系统仿真模型。结果表明：达到同样的校正效果时,采用组合优化的算法收敛速度要明显优于基于区域法的收敛速度,从而验证了理论分析的合理性。     

自适应光学系统随机并行梯度下降控制算法实验研究

随机并行梯度下降算法是一种极具应用潜力的自适应光学系统控制算法,具有不依赖波前传感器直接对系统性能指标进行优化的特点。基于32单元变形镜、CCD成像器件等建立自适应光学系统随机并行梯度下降控制算法实验平台。考察算法增益系数和扰动幅度对校正效果和收敛速度的影响,验证随机并行梯度下降算法的基本原理。实验结果表明参量选取合适的情况下,随机并行梯度下降控制算法对静态或慢变化的畸变波前具有较好的校正能力。根据实验结果分析了影响随机并行梯度下降算法校正速度的主要因素。     

自适应光学系统随机并行梯度下降控制算法仿真与分析

随机并行梯度下降算法能不依赖波前传感器直接对系统性能进行优化。以32单元变形镜为校正器,采用随机并行梯度下降算法建立了自适应光学系统仿真模型。通过分析该系统对静态波前畸变的校正能力,验证了随机并行梯度下降算法的收敛性;讨论了算法增益系数、随机扰动幅度与收敛速度的关系,并指出通过算法增益系数的自适应调整可以改进算法的收敛速度。     

自适应光学系统随机并行梯度下降算法

随机并行梯度下降(SPGD)算法可以对系统性能指标直接优化来校正畸变波前。对基于SPGD算法的61单元自适应光学系统进行仿真模拟,分析了对不同初始静态畸变波前的校正能力,并比较了不同性能指标情况下的算法增益系数、扰动幅度值的选取及校正情况。仿真结果表明:算法收敛速度很大程度上依赖于增益系数和扰动幅度值,对畸变较大的波前,随机扰动幅度在0.50~0.85范围内,性能指标采用焦斑平均半径比采用斯特列尔比取得的校正效果好。     

基于Zernike模式的随机并行梯度下降算法的收敛速率

为了加快控制变形镜进行波前整形的随机并行梯度下降（SPGD）算法的收敛速率,提高实时波前整形能力,本文利用由12阶Zernike多项式构成的畸变波前和32单元变形镜建立了仿真模型。基于Zernike多项式的单位正交性,得到了两个常数矩阵,当斯特列尔比（SR）达到0.8时,需要算法迭代660次,简化了算法的运算过程,加快了算法运行时间。通过Matlab7.8.0对6种SPGD算法进行仿真对比,结果显示：当SR要求不高时,可使用间接固定双边SPGD算法来提高收敛速度;当SR要求较高时,则应当使用间接自动双边SPGD算法。提出的算法为实际的激光整形提供了理论指导。     

随机并行梯度下降光束净化实验研究

利用自适应光学技术进行光束净化是高能激光系统中一项重要的研究内容.为实现光束净化系统的小型化和低成本,基于系统性能评价函数无模型最优化的波前畸变校正方法是适合的技术方案.就随机并行梯度下降(SPGD)最优化算法在光束净化系统中的应用展开研究.针对高能激光束常见的像差分布进行了SPGD波前校正的数值模拟,在此基础上构建了37单元自适应光学光束净化实验平台,讨论了双边扰动梯度估计和迭代增益系数自适应变化对算法收敛特性的影响.数值模拟与实验结果验证了SPGD算法对不同程度波前畸变的校正能力,表明了SPGD光束净化方案的可行性.     

随机并行梯度下降算法性能与变形镜排布规律的关系研究

对随机并行梯度下降算法(SPGD)性能与不同变形镜排布规律的关系进行了研究。以采用Roddier方法生成的由52项Zernike像差构成的畸变波前为整形对象,对SPGD算法的收敛速率和整形效果与变形镜排布规律(单元数分别为19、21、32、37、45、60、61、77、91)之间的关系进行了仿真研究。结果表明:从整体分析,随着变形镜单元数逐渐增多,SPGD算法的收敛速率和整形效果均逐渐变差;从局部分析,由于变形镜元胞类型变化和边缘占空比的影响,在渐变规律中产生了局部差异。     

Bandwidth analysis and improvement of the beam phasing of fiber amplifiers using the stochastic parallel gradient descent algorithm

We present the numerical analysis and improvement of the bandwidth of the beam phasing of fiber amplifiers using the stochastic parallel gradient descent (SPGD) algorithm. Time-varying phase distortions are incorporated into the dynamic simulation. It is shown by numerical calculation that the bandwidth of the standard SPGD algorithm is dependent on the phase distortions and on the number of lasers. The time-averaged cost function will decrease with an increase in the amplitude and frequency of phase distortions, and in the number of lasers. A cross-grouped SPGD algorithm is proposed, which offers attractive performance due to its robustness to various time-varying phase distortions and faster convergence rate for beam phasing of larger number of lasers. Dynamic simulation shows that the control bandwidth can be improved using the cross-grouped SPGD algorithm.     

Adaptive optics correction based on stochastic parallel gradient descent technique under various atmospheric scintillation conditions: numerical simulation

An adaptive optics system utilizing a Shack–Hartmann wavefront sensor and a deformable mirror can successfully correct a distorted wavefront by the conjugation principle. However, if a wave propagates over such a path that scintillation is not negligible, the appearance of branch points makes least-squares reconstruction fail to estimate the wavefront effectively. An adaptive optics technique based on the stochastic parallel gradient descent (SPGD) control algorithm is an alternative approach which does not need wavefront information but optimizes the performance metric directly. Performance was evaluated by simulating a SPGD control system and conventional adaptive correction with least-squares reconstruction in the context of a laser beam projection system. We also examined the relative performance of coping with branch points by the SPGD technique through an example. All studies were carried out under the conditions of assuming the systems have noise-free measurements and infinite time control bandwidth. Results indicate that the SPGD adaptive system always performs better than the system based on the least-squares wavefront reconstruction technique in the presence of relatively serious intensity scintillations. The reason is that the SPGD adaptive system has the ability of compensating a discontinuous phase, although the phase is not detected and reconstructed.     

Adaptive correction of vortex laser beam in a closed-loop system with phase only liquid crystal spatial light modulator

We propose and demonstrate the wave front correction of a vortex laser beam by using dual phase only liquid crystal spatial light modulators (LC-SLMs) and a stochastic parallel gradient descent (SPGD) algorithm. One phase only LC-SLM is used to generate vortex laser beam by loading spiral phase screen onto the wave front of input quasi-Gaussian beam. The other phase only LC-SLM under SPGD controller based on the subzone control method adaptively compensates the wave front of vortex laser beam. Numerical simulation and experimental results show that after correction, vortex doughnut like beam is focused into a beam with airy disk pattern distribution in the far field. The adaptive corrections of vortex laser beam with different optical topological charges are studied. The results show that the optical topological charge has little influence on adaptive correction. The powers in the main lobe of far field intensity distributions of vortex laser beams with different optical topological charges are all greatly improved by adaptive correction. The technique proposed in this paper can be used in optical communication, relay mirror and atmospheric turbulence correction.     

自适应光学随机并行梯度下降算法波前整形规律仿真

本文首先介绍了基于Zernike模式的SPGD算法对大气湍流畸变波前的整形原理,通过推导得到了关于性能指标的简明表达式,使SPGD算法收敛速率得到明显提升。然后建立了自适应光学随机并行梯度下降算法波前整形系统模型,主要对SPGD算法收敛速率、整形能力和整形效果随波前畸变量和变形镜模型的变化规律作了较为详细的仿真研究,整体定性结果表明：三者的变化规律有一定的相似性,同时利用最小二乘法得到了关于整形能力和整形效果变化规律的定量表达式,若从自适应光学波前整形系统的实时性和简单性考虑,在保证一定整形效果的情况下,选择37单元变形镜对畸变波前的3~27（25）阶Zernike像差进行整形即可。     

自适应光学中SPGD算法关键参数实时调节方法

介绍了随机并行梯度下降（SPGD）算法及其在相干合成中的应用,针对实验中算法关键参数难以调节的难点,提出采用软硬件结合的新方式,实现对实验数据的在线采集和分析以及对SPGD算法关键参数的自动实时调节。开展了4路光纤激光相干合成实验,对不同调节方法进行对比。实验中采用新方式有效调节了SPGD算法中增益系数和随机扰动幅度的取值,合成效果显著。     

基于分区域耦合的随机并行梯度下降自适应光学

基于随机并行梯度下降(SPGD)方法的自适应光学(AO)系统通过直接优化系统的性能评价函数来控制波前校正器以补偿光束中存在的波前畸变。为了提高这种无模型优化自适应光学系统的收敛速度, 提出了基于分区域耦合的新方法以改进传统随机并行梯度下降自适应光学系统的工作方式。将波前校正器光学孔径分成多块子区域, 每块子区域对应着的所有驱动器作为一个整体控制单元, 从形式上可以得到一个空间分辨率较低的分区域波前校正器。该校正器与原校正器同步工作, 并采用随机并行梯度下降算法对同一个性能评价函数进行优化, 从而构成了双校正器的耦合工作结构。对256单元分立活塞式波前校正器建立了自适应成像系统的数值模型, 结果表明这种分区域耦合的随机并行梯度下降自适应光学系统比传统随机并行梯度下降自适应光学系统具有更快的收敛速度和更好的渐近态。     

Improving performance of FxRLS algorithm for active noise control of impulsive noise

Active noise control (ANC) systems employing adaptive filters suffer from stability issues in the presence of impulsive noise. New impulsive noise control algorithms based on filtered-x recursive least square (FxRLS) algorithm are presented. The FxRLS algorithm gives better convergence than the filtered-x least mean square (FxLMS) algorithm and its variants but lacks robustness in the presence of high impulsive noise. In order to improve the robustness of FxRLS algorithm for ANC of impulsive noise, two modifications are suggested. First proposed modification clips the reference and error signals while, the second modification incorporates energy of the error signal in the gain of FxRLS (MGFxRLS) algorithm. The results demonstrate improved stability and robustness of proposed modifications in the FxRLS algorithm. However, another limitation associated with the FxRLS algorithm is its computationally complex nature. In order to reduce the computational load, a hybrid algorithm based on proposed MGFxRLS and normalized step size FxLMS (NSS-FXLMS) is also developed in this paper. The proposed hybrid algorithm combines the stability of NSS-FxLMS algorithm with the fast convergence speed of the proposed MGFxRLS algorithm. The results of the proposed hybrid algorithm prove that its convergence speed is faster than that of NSS-FxLMS algorithm with computational complexity lesser than that of FxRLS algorithm.     

Optimal nonmonotonic convergence of the iterative Fourier-transform algorithm

The increase of the monotonic convergence rate is an important issue for iterative Fourier-transform algorithms. However, the steepest monotonic convergence of the iterative Fourier-transform algorithm does not always promise an optimal solution in the design of a diffractive optical element. The optimal nonmonotonic convergence of the iterative Fourier-transform algorithm is investigated by employing a microgenetic algorithm. The proposed hybrid scheme of the iterative Fourier-transform algorithm and the microgenetic algorithm show nonmonotonic convergence, and this results in a superior design.     

An enhanced non-uniformity correction algorithm for IRFPA based on neural network

Influenced by detector materials’ non-uniformity, growth and etching techniques, etc., every detector’s responsivity of infrared focal plane arrays (IRFPA) is different, which results in non-uniformity of IRFPA. And non-uniformity of IRFPA generates fixed pattern noises (FPN) that are superposed on infrared image. And it may degrade the infrared image quality, which greatly limits the application of IRFPA. Non-uniformity correction (NUC) is an important technique for IRFPA. The traditional non-uniformity correction algorithm based on neural network and its modified algorithms are analyzed in this paper. And a new improved non-uniformity correction algorithm based on neural network is proposed in this paper. In this algorithm, the desired image is estimated by using three successive images in an infrared sequence. And blurring effect caused by motion is avoided by applying implicit motion detection and edge detection. So the estimation image is closer to real image than the estimation image estimated by other algorithms, which results in fast convergence speed of correction parameters. A comparison is made to these algorithms in this paper. And experimental results show that the algorithm proposed in this paper can correct the non-uniformity of IRFPA effectively and it prevails over other algorithms based on neural network.     

Correcting dynamic residual aberrations of conformal optical systems using AO technology

This paper analyses the dynamic residual aberrations of a conformal optical system and introduces adaptive optics (AO) correction technology to this system. The image sharpening AO system is chosen as the correction scheme. Communication between MATLAB and Code V is established via ActiveX technique in computer simulation. The SPGD algorithm is operated at seven zoom positions to calculate the optimized surface shape of the deformable mirror. After comparison of performance of the corrected system with the baseline system, AO technology is proved to be a good way of correcting the dynamic residual aberration in conformal optical design.     

基于随机并行梯度下降算法光纤激光相干合成的高精度相位控制系统

随机并行梯度下降(SPGD)算法是实现大功率主振荡功率放大器相干合成的有效手段之一.分析了SPGD算法用于相干合成的基本原理,设计了基于SPGD算法的相位控制系统,给出了基于该SPGD控制器的相干合成实验结果.理论分析表明,SPGD算法控制器的迭代速率为200kHz,对于两路相干合成的平均控制带宽大于12.5kHz,最优控制精度可达1/179个波长.实验表明,SPGD控制器能够实现高速的相位控制.在相位噪声幅度大于1/10个波长、频率为3kHz的情况下,有效地实现了两路激光的相位锁定,相位残差控制在1/25个波长内.