Adaptive optics for airborne platforms—Part 2: controller design

In an airborne platform adaptive optics application, the inbound (beacon) and outbound (high-energy laser) wavefronts propagate through different regions of the atmosphere at different time instants, that is, spatial and temporal anisoplanatism cannot be neglected. Measurements in an airborne platform system are from the inbound (beacon) wavefront and therefore, the outbound, high-energy laser's wavefront phase distortion Zernike expansion coefficients must be estimated. Once the said estimates are available, these values are used by a linear quadratic regulator to drive the actuators of the deformable mirror. The controller, which consists of a Kalman filter estimator in tandem with the regulator, provides commands to the piezoelectric actuators of the deformable mirror. Thus, the estimated conjugate phase is applied to the mirror and, hence, to the outbound high-energy laser wavefront, such that at the aim point on the target, the high-energy laser wavefront distortion is minimized. In other words, the high-energy laser is correctly pointed to the aim point and the Strehl ratio is maximized. In this way, the correct deformation is applied to the deformable mirror and the benefits of adaptive optics are realized in an airborne platform application. In Part 2 of this paper, the design of the controller, that is, a Kalman filter and regulator, is addressed. The theoretical derivations are validated in extensive simulation experiments.     

Limitations of the Shack–Hartmann sensor for testing optical aspherics

Aspheric lenses and surfaces are increasingly used in modern high-quality optics. Therefore, new measuring methods for an accurate quantification of these aspheres are also necessary. The current approach to quantify aspheres is to apply null systems such as computer-generated holograms as a part of a null lens in a interferometer. An alternative to this method is the Shack–Hartmann wavefront sensor. The dynamic range of this sensor can be adjusted by the optical parameters of the applied microlens array. Hence, large wavefront aberrations can be measured directly without a null lens. However, there are basic limitations in the dynamic range of a Shack–Hartmann sensor (SHS) depending on the curvature of the incident wavefront. In this paper, an analytical expression to determine the strongest wavefront curvature which can be measured with a defined microlens array of an SHS is derived. It allows to calculate the microlens parameters required to measure the wavefront of a test lens. Particularly, the influence of rotational symmetric aspherical wavefront shapes to the dynamic range of an SHS has been studied. A comparison between interferometry and the SHS has been accomplished. Numerical solutions using scalar diffraction theory illustrate the analytical predictions.     

神光II升级装置波前传输计算模型与验证

基于衍射理论和神光II升级激光装置的光路结构及光学元件波前像差,建立了激光装置光束波前像差传输计算模型,用以获取在实验中难以取得,但对激光系统设计和安全运行非常重要的变形镜校正面型、放大器动态波前像差、空间滤波器波前像差和焦斑形态等过程数据。利用该模型对激光装置常规运行时的静态传输、变形镜加压静态传输和大能量发射3种状态下的波前传输进行理论分析,计算结果与实验测量相一致,并在此基础上对变形镜未加压大能量发射波前传输进行计算。计算模型可用于指导相关器件的设计,优化激光运行控制模式,有利于提高激光装置输出波前质量。     

大口径光学元件中频波前的检测

大口径光学元件中频波前的准确评价已成为高功率激光系统中关注的焦点,元件中频波前均方根值是重要评价指标之一。根据波前中频检测频段及波前检测设备频响特性,将波前的中频区域分为两个检测频段,分别采用干涉仪和光学轮廓仪实现了中频波前均方根值的检测。采用大口径干涉仪可实现全口径波前中频区域低频段波前的检测,通过比对大口径干涉仪和采用小口径干涉仪结合分块融合平均方法的检测结果,提出采用分块融合平均方法也可检测相应频段全口径波前均方根。采用光学轮廓仪通过离散采样的方法检测大口径元件中频区域高频段波前均方根,针对不同离散采样方式的实验结果表明:33的采样方式能满足对410 mm410 mm口径元件中频区域高频段波前均方根的检测。     

固体板条MOPA激光光束质量主动控制

详细介绍了固体板条MOPA激光系统输出光束特性,包括光束近场强度分布、波前畸变空间特性和时间频率特性等,并以此为设计输入,研制了大动态范围高空间分辨率变形镜。研制的自适应光学闭环校正系统在强光条件下,实现了全程闭环控制,光束质量因子平均值从开环7.4提高到4.06。根据当前波前闭环校正残差,提出了进一步提高MOPA系统输出光束质量的方法。     

用于激光二极管抽运固体激光器热畸变补偿的微变形镜特性研究

在很多高功率的激光器中,由于热效应的缘故会导致激光器的光束质量大幅下降。为了改善激光光束的动态畸变,自适应光学是一种有效的方法。作为自适应光学系统的核心部件,设计了一种可用于大功率激光二极管抽运固体激光器(DPL)热畸变补偿的新型可变形反射镜,并且通过工艺流片实际制造出有效反射面积为30 mm×30 mm,拥有49个静电驱动单元的可变形反射镜。对该微机电系统(MEMS)微变形镜的变形特性进行了详细的理论分析和模拟研究,并和实际测量结果进行了比较,得到了满意的结果。为了实现闭环控制,对微变形反射镜的光学影响函数矩阵进行了全面测量,并用得到的电压矩阵对畸变激光光束的波前进行了校正。实验结果表明,微变形反射镜可以有效地改善激光光束质量。     

二元光学在激光测距中的应用

利用二元光学器件的相位补偿作用进行波面变换,将高斯光束变换为平面光束,达到减小激光测距仪光束发散角的目的。在２０ｋｍ 处对激光光束发散角采用叠加积分衍射分析进行计算,得到结果为２ ×１０ － ４ｍｒａｄ , 远小于原来的２ｍｒａｄ 。计算还表明测程没有因能量损失而降低。     

Fast wavefront sensing using a hardware parallel classifier chip

In most applications of laser technology and optics the beam quality, the ability to focus a laser beam and the achievement of a good optical resolution play an important role. The compensation of distortions using adaptive optics requires fast wavefront measurement. Classical wavefront analysis schemes use matrix operations, which show a nonlinear computation time dependence with matrix size, making it difficult to achieve high-speed control loops at a high resolution. A novel wavefront sensor system is presented using a massively parallel k-nearest neighbor classifier chip in an embedded hardware setup. Our miniaturized sensor is able to detect one optical distortion within about 80 μs allowing its use for high-speed adaptive optics applications.     

用蒙特卡罗方法计算棒状放大器内的泵浦能量分布

建立了用蒙特卡罗和光线跟踪法计算高功率固体激光系统棒状放大器能量沉积分布的模拟程序。在本文的模型中正确考虑了光线在反射器激光棒上的多次反射，氙灯的辐射光谱以及钕玻璃的吸收光谱。计算结果与作者用阈值法测量的增益均匀性以及用哈特曼网格法测量的氙灯泵浦对激光波面畸变的影响是一致的。     

高功率固体激光装置哈特曼传感器参考波前标定方法

 在高功率固体激光驱动器中,参考波前标定是自适应光学系统应用过程中的关键技术之一。基于国内某大型激光原型装置光路特点,对比研究了两种不同的参考波前标定方法,详细分析了两种方法的标定过程和影响标定结果准确性的主要因素,选取了较为准确的参考波前标定结果。基于该参考波前,对AO系统闭环和开环条件下的远场进行了对比测试,实验证明了标定结果的有效性。该研究结果成功应用于该激光原型装置的AO系统波前补偿实验。     

基于瀑布型多重网格加速的复指数波前复原算法

激光在大气中传输时,由于强湍流或长传输距离的影响,畸变波前中出现由相位起点组成的不连续相位,现有波前复原算法不能有效复原不连续相位,使得自适应光学系统校正效果下降甚至失效.本文分析了最小二乘波前复原算法不能复原相位奇点的原因,提出了基于瀑布型多重网格加速的复指数波前复原算法,给出了复指数波前复原算法中迭代计算、降采样、插值计算的实现方式.研究了该方法对不连续相位和随机连续相位的复原能力,数值分析了采用复指数波前复原算法的自适应光学系统对大气湍流像差的校正效果.仿真结果表明,同等复原精度下,相比直接迭代过程,该方法所需浮点乘数目减少了近2个数量级,且随着夏克-哈特曼波前传感器子孔径数目增加,其在计算量上的优势更加明显. Rytov方差较大时,相比直接斜率法,自适应光学系统采用复指数波前复原算法后校正光束Strehl比提升1倍.     

基于双向反射函数的水下激光引信回波仿真方法

针对鱼雷激光近炸引信探测水下近场目标的需求,开展了水下激光引信回波蒙特卡洛仿真方法研究.结合水下激光引信探测特点建立水下目标回波的蒙特卡洛仿真模型.为了提高水中非朗伯目标表面回波仿真的准确度,推导了基于双向反射函数的光子反射方向概率分布,根据概率分布随机抽样光子反射方向.仿真了不同距离和入射角度条件下的水中目标回波信号.仿真结果表明:目标回波幅度随目标距离和入射角度的增大迅速下降,目标距离在6~12m内变化时,信号峰值动态范围为11.5dB;目标距离为8m,激光入射角在0~45°内变化时,信号峰值动态范围为9.2dB.为验证仿真方法的正确性,在水池中进行水中目标蓝绿激光探测实验,实验结果和仿真结果一致.研究成果可为解决传统蒙特卡洛方法在水中非朗伯面目标回波仿真中的适用性问题及水下激光引信优化设计提供参考.     

用于活体人眼视网膜观察的自适应光学成像系统

利用自适应光学技术,研制了两套活体人眼视网膜高分辨力成像系统,在实时校正人眼波前误差的基础上,实现活体人眼视网膜细胞尺度的高分辨力成像。这两套系统分别采用19和37单元小型压电变形反射镜作为波前校正元件,哈特曼-夏克(Hartmann-Shack)波前传感器测量波前误差,用眼底反射的半导体激光作为波前探测的信标。在用计算机控制自适应光学系统实现人眼波前误差校正后,触发闪光灯照明视网膜,用CCD相机记录视网膜的高分辨力图像。校正后的残余波前误差的均方根值已分别小于1／6和1／10波长,相当于视网膜上成像分辨力分别为3．4μm和2．6μm,接近衍射极限。试验表明37单元系统的成像质量更好。     

Liquid deformable mirror for high-order wavefront correction

We propose and demonstrate a novel liquid deformable mirror, based on electrocapillary actuation, for high-order wavefront correction. The device consists of a two-dimensional array of vertically oriented microchannels filled with two immiscible liquids, an aqueous electrolyte, and a viscous dielectric liquid, where the dielectric liquid overfills the top end of the channel and forms a thin layer on top. To remedy the poor reflectivity of pure liquids, a free-floating reflective membrane or a dye-coated liquid can be used. The proposed device offers several advantages for adaptive optics applications. These advantages include a high number of actuators, high stroke dynamic range, low power dissipation, fast response time, an initially flat surface, and low cost. However, the device is mainly suitable for dynamic wavefront correction and is limited by its orientation.     

基于深度学习的自适应光学波前传感技术研究综述

波前传感是自适应光学系统的重要组成部分,在地基大口径望远镜、激光大气传输、无线光通信、激光驱动核聚变等领域发挥了关键作用,同时也常应用于自由曲面的光学测量中。与此同时,深度学习作为一种较为通用的前沿技术,成功在计算机视觉、自然语言处理等众多领域取得了革命性进展。使用深度学习的方法改进自适应光学系统中的波前传感器,以期实现更精准的波前探测,以及适应更复杂的应用场景是自适应光学的发展趋势,也是深度学习应用领域的一个新课题。介绍了深度学习在自适应光学波前传感中的应用现状,主要分析了在相位反演波前传感器和哈特曼波前传感器中的研究特点,并在最后进行了总结和展望。     

高功率激光装置中局部波前畸变的非线性传输

 根据高功率激光装置中强激光束传输的特点，建立了描述光束通过光学元件时引入的局部波前畸变模型，并利用纹波传输的线性化处理方法，研究了带有局部波前畸变的高强度光束的传输规律。以一个连续的非增益激光介质为例，用该模型进行数值模拟，给出了不同空间尺度的局部波前畸变的频谱分布、不同空间频率的纹波引起的振幅非线性增长曲线、不同B积分的非线性增益随光束传输距离的变化曲线,光束振幅非线性增益达到最大时光束的传输距离和标准传输距离不同时由局部波前畸变引入的振幅分布。研究表明为了防止光学元件损坏，应避免光学元件表面出现半径为0.5~2.0 mm的局部瑕疵。     

A new Monte Carlo code for absorption simulation of laser-skin tissue interaction

In laser clinical applications, the process of photon absorption and thermal energy diffusion in the target tissue and its surrounding tissue during laser irradiation are crucial. Such information allows the selection of proper operating parameters such as laser power, and exposure time for optimal therapeutic. The Monte Carlo method is a useful tool for studying laser-tissue interaction and simulation of energy absorption in tissue during laser irradiation. We use the principles of this technique and write a new code with MATLAB 6.5, and then validate it against Monte Carlo multi layer (MCML) code. The new code is proved to be with good accuracy. It can be used to calculate the total power bsorbed in the region of interest. This can be combined for heat modelling with other computerized programs.     

Wavefront reconstruction of high power laser diode array

A new method for the wavefront measurement of high power laser diode array with large dynamic ranges is presented. The wavefront sensor has a large dynamic range of −π/2 to π/2 and high precision on the measured wavefront slopes (±1′). According to the measurements of the slopes and intensity of the high power laser diode array, complex amplitudes can be obtained and the wavefront can be reconstructed. The reconstructed wavefront of high power laser diode array can be used in designing a laser beam shaping system in future work.     

Scaling of a collapsed polymer globule in two dimensions

Extensive Monte Carlo data analysis gives clear evidence that collapsed linear polymers in two dimensions fall in the universality class of athermal, dense self-avoiding walks, as conjectured by Duplantier [Phys. Rev. Lett. 71, 4274 (1993)].10.1103/PhysRevLett.71.4274 However, the boundary of the globule has self-affine roughness and does not determine the anticipated nonzero topological boundary contribution to entropic exponents. Scaling corrections are due to subleading contributions to the partition function corresponding to polymer configurations with one end located on the globule-solvent interface.     

Wavefront response matrix for closed-loop adaptive optics system based on non-modulation pyramid wavefront sensor

Pyramid wavefront sensor (PWFS) is a kind of wavefront sensor with high spatial resolution and high energy utilization. In this paper an adaptive optics system with PWFS as wavefront sensor and liquid-crystal spatial light modulator (LC-SLM) as wavefront corrector is built in the laboratory. The wavefront response matrix is a key element in the close-loop operation. It can be obtained by measuring the real response to given aberrations, which is easily contaminated by noise and influenced by the inherent aberration in the optical system. A kind of analytic solution of response matrix is proposed, with which numerical simulation and experiment are also implemented to verify the performance of closed-loop correction of static aberration based on linear reconstruction theory. Results show that this AO system with the proposed matrix can work steadily in closed-loop operation.