Adaptive wavefront control with asynchronous stochastic parallel gradient descent clusters

A scalable adaptive optics (AO) control system architecture composed of asynchronous control clusters based on the stochastic parallel gradient descent (SPGD) optimization technique is discussed. It is shown that subdivision of the control channels into asynchronous SPGD clusters improves the AO system performance by better utilizing individual and/or group characteristics of adaptive system components. Results of numerical simulations are presented for two different adaptive receiver systems based on asynchronous SPGD clusters-one with a single deformable mirror with Zernike response functions and a second with tip-tilt and segmented wavefront correctors. We also discuss adaptive wavefront control based on asynchronous parallel optimization of several local performance metrics-a control architecture referred to as distributed adaptive optics (DAO). Analysis of the DAO system architecture demonstrated the potential for significant increase of the adaptation process convergence rate that occurs due to partial decoupling of the system control clusters optimizing individual performance metrics.     

Decoupled stochastic parallel gradient descent optimization for adaptive optics: integrated approach for wave-front sensor information fusion

A new adaptive wave-front control technique and system architectures that offer fast adaptation convergence even for high-resolution adaptive optics is described. This technique is referred to as decoupled stochastic parallel gradient descent (D-SPGD). D-SPGD is based on stochastic parallel gradient descent optimization of performance metrics that depend on wave-front sensor data. The fast convergence rate is achieved through partial decoupling of the adaptive system's control channels by incorporating spatially distributed information from a wave-front sensor into the model-free optimization technique. D-SPGD wave-front phase control can be applied to a general class of adaptive optical systems. The efficiency of this approach is analyzed numerically by considering compensation of atmospheric-turbulence-induced phase distortions with use of both low-resolution (127 control channels) and high-resolution (256 x 256 control channels) adaptive systems. Results demonstrate that phase distortion compensation can be achieved during only 10-20 iterations. The efficiency of adaptive wave-front correction with D-SPGD is practically independent of system resolution.     

Optimizing stochastic gradient descent algorithms for serially addressed adaptive-optics wavefront modulators

High-resolution adaptive-optical systems with thousands to millions of pixels will most likely have to employ serial- or matrix-addressed spatial light modulators (e.g., microelectromechanical-system-on-VLSI spatial light modulators). We compare parallel gradient descent adaptive-optics algorithms with serial gradient descent algorithms running on serially addressed modulators. While serial algorithms have previously been shown to require more iterations than parallel algorithms, we show that, because of the limitations of the databus, each serial iteration of the algorithm on a serial modulator requires significantly less time to complete than a parallel iteration, thereby favoring the serial algorithm when time to convergence is used as the performance metric. Thus, such high-resolution serially addressed devices are generally better matched to the serial-update wavefront correction algorithm owing to the data load penalty imposed by the bandwidth-limited databus of these modulators.     

A new parallel sparse direct solver: Presentation and numerical experiments in large‐scale structural mechanics parallel computing

The main purpose of this work is to present a new parallel direct solver: Dissection solver. It is based on LU factorization of the sparse matrix of the linear system and allows to detect automatically and handle properly the zero‐energy modes, which are important when dealing with DDM. A performance evaluation and comparisons with other direct solvers (MUMPS, DSCPACK) are also given for both sequential and parallel computations. Results of numerical experiments with a two‐level parallelization of large‐scale structural analysis problems are also presented: FETI is used for the global problem parallelization and Dissection for the local multithreading. In this framework, the largest problem we have solved is of an elastic solid composed of 400 subdomains running on 400 computation nodes (3200 cores) and containing about 165 millions dof. The computation of one single iteration consumes less than 20 min of CPU time. Several comparisons to MUMPS are given for the numerical computation of large‐scale linear systems on a massively parallel cluster: performances and weaknesses of this new solver are highlighted. Copyright © 2011 John Wiley & Sons, Ltd.     

激光成丝不稳定性并行数值模拟研究

实现了三维非线性流体力学与激光传播耦合的并行数值模拟程序,用来研究高强度激光在稀疏等离子体中的激光成丝不稳定性.并行数值实验结果表明,该程序在64台处理机上取得了较好的并行加速比,且能定性模拟激光成丝现象.     

温度梯度场中梁的形变控制研究

5介绍了一种新型的压电作动器—层叠式压电作动器,并使用这种作动器对温度场中的梁进行了形状控制研究。根据哈密顿原理,得到了粘贴有层叠式压电作动器的梁结构的控制方程,进行了数值仿真,并且用Comsol软件进行了模拟,两者的结果基本一致。对压电作动器的控制电压进行了优化,得到了最优控制电压。由于层叠式压电作动器的控制力与压电片的层数成二次函数关系,当控制电压恒定时,层叠式压电作动器的控制力随着压电器层数的增加而迅速减小。使用层叠式压电作动器可以在比其他作动器更小的电压下取得更好的控制效果。通过与普通压电作动器的比较,可以发现层叠式压电作动器可以有效地降低作动器的施加电压,而且可以显著增强控制效果。这种形状控制方法为应用层叠式压电作动器进行薄壁结构的形状控制提供了理论基础。     

Electromagnetic control of convective heat transfer during electron beam evaporation: model experiments

The present paper aims to demonstrate that melt-flow during electron beam evaporation can be effectively controlled by using external magnetic fields to considerably reduce the convective heat transfer. We discuss the various effects of a static magnetic field, a static field combined with an applied electrical current, and a rotating magnetic field. We perform model experiments using GaInSn in eutectic composition as a test liquid. The liquid metal is heated locally at its free surface by an electric resistance heater. The results of the measurements are compared to prediction of numerical simulations.     

Pulsed laser beam control by hot spot tracking

The use of a self-developed body-centered reference for precision pulsed laser hot spot tracking is examined through a numerical computer simulation based on a solution of the heat equation with time varying input flux. Tracking resolution and sensitivity in two commonly used infrared sensor bands are compared. To cope with inherent instability, a simple compensation scheme is suggested with a demonstrated improvement in boresight error drift. Optimum performance is obtainable by judicious selection of sensor wavelength bands and sampling times relative to laser pulse time.     

High-quality parallel patterning of carbon nanotube thin films by a pulsed laser beam

Single-walled carbon nanotube thin films solution-deposited on the glass substrate were directly patterned by a spatially-modulated pulsed laser beam (wavelength = 355 nm, pulse width = 5 ns) incident from the backside of the substrate. This method utilizes a ultrashort pulse-induced strong thermo-elastic force exerting on the film which plays a role to detach it from the substrate. The threshold energy density required for patterning was as low as 90 mJ/cm², making it possible to pattern over a few square centimeters by a single pulse with maximum energy of 180 mJ. The irradiated regions of the film were clearly photoetched without leaving any residual nanotubes. High-fidelity patterns could be fabricated with a feature size of 35 μm.     

Phase control requirements of high intensity laser beam combining

Aiming at getting the general requirements of the beam combine for ignition scale laser facilities, the analytical expressions including the factors affecting the combine results are derived. The physical meanings of every part are illustrated. Based on these expressions, the effects of the factors, including the beam configuration, piston error, and tip/tilt error, are studied analytically and numerically. The results show that the beam configuration cannot affect the Strehl ratio (SR) of the combined beam, but it influences the FWHM of the main peak and the ratio of the main peak and the side peak. The beam separation should be no more than 1.24 times the individual beam width for the multibeam combine, and be close to the individual beam width for the two-beam combine as much as possible. The piston error can change the characteristics of the combine beam focus, including the peak intensity, the focal spot morphology, the fractional energy contained within a certain area, and the center of mass. For the two-beam combine, a piston error less than 2π/5 rad is suitable, and for the multibeam combine, the standard deviation of the piston error should be no more than 2π/10 rad. The tip/tilt error has a great influence on the combined results. It affects the superposition degree of the focal spots of the combined elements directly. A requirement of 0.5~1 μrad for the standard deviation of the tip/tilt error is adequate.     

Fabrication of gradient material by electron beam smelting based on scanning track control

A new electron beam (EB) control system was developed in a general vacuum EB machine by equipping it with an industrial control computer, programmable logic control (PLC), deflection coil, data acquisition card, power amplifier, etc. In this control system, the scanning track and energy distribution of the EB could be edited off-line, adjusted in real-time, and controlled on-line. Ti-Mo gradient material (GM) with high temperature resistance was fabricated using electron beam smelting (EBS) control. The smelting pro cesses include three steps such as preheating, smelting, and homogenizing. The results show that GM prepared by using smelting technology has fine appearance, and has good integrated interface with Ti alloy. The Mo and Ti elements are gradual diversification in the interface of the gradient material. The microstructure near the Ti alloy base metal is α+β basket-waver grain, and the microstructure near GM is single phase of β solid solution.     

Scattering of SAW at discontinuities: some numerical experiments

In developing models to characterize the metallic-finger step discontinuity in SAW devices on piezoelectric substrates it is important to estimate the SAW reflection coefficients and the effect of other modes at step discontinuities. In the literature perturbation approximations, finite element methods and various phenomenological parameter-fitting using experimental data are among the techniques which have been exploited. In this paper, a nonperturbation approximation method, which attempts to parallel some well-established techniques used in solving electromagnetic waveguide-discontinuity problems, is explored as an alternate numerical procedure for calculating SAW reflections. Implementing such a method requires that the full solution to SAW propagation in the free and in the layered regions be obtained. Thus, for a free-to-metallized discontinuity, the two SAW phase velocities, the total SAW power now in the free region and in the layered region, and the electroacoustic fields, must be calculated. In this paper the reflection problem is formulated as an optimization problem subject to the necessary scattering parameter constraints. Numerical experiments are described, and the results of calculations for reflectivity are compared to perturbation approximations, nonperturbation approximations and the available experimental data for the effects of energy storage     

Two-dimensional bioluminescence tomography: numerical simulations and phantom experiments

The reconstruction of internal light sources in bioluminescence tomography (BLT) is a challenging inverse problem because of the limited amount of information available compared with that for other kinds of tomography such as fluorescence tomography in which external illumination sources are used. We demonstrated previously, using phantom experiments, that a target containing luciferases could be detected tomographically when the target was located relatively close to the imaging boundary. Here we describe an improved BLT reconstruction method that can detect luciferase-containing targets located anywhere within an imaging domain. The method is tested with numerical simulations and further confirmed with several phantom experiments.     

准分子激光前端光束形态控制技术研究

在高功率准分子激光系统中,前端光束形态直接决定着系统输出的光束形态,并直接与靶物理需求密切相关.主要介绍了窄脉冲准分子激光前端光束形态控制的实验研究进展,基于散射法获得了满足系统要求的部分相干源,直接利用三束激光脉冲堆积获得了平顶整形脉冲,并结合放大实验结果对上述实现方法进行了评价分析.     

Adaptive optics compensation for propagation through deep turbulence: initial investigation of gradient descent tomography

To implement adaptive optics compensation for propagation through deep turbulence, the concept of gradient descent tomography has been developed. Here two or more deformable mirrors are controlled by an efficient iterative algorithm that optimizes the integral I(2) image-sharpening metric. In this work a difficult case involving imaging over a 2 km path with a C(n)(2) of 2 x 10(-13)m(-2/3) is considered. For a wavelength of 1.06 microm and a 10-cm-diameter aperture, lambda/D is seven times the isoplanatic angle (theta(0)=1.54 microrad), and the Rytov number is 5.5. For three points placed along a line spanning approximately 70 isoplanatic patch sizes all three points are compensated somewhat, illustrating that anisoplanatism is addressed. The fact that the corresponding performance improvement ratios are 1.20, 1.34, and 3.26 in the presence of such strong scintillation and anisoplanatism is quite significant.     

Optimum control of the laser beam intensity profile with a deformable mirror

A new deformable mirror control system is developed. This system consists of a deformable mirror, a CCD camera, an image processor, a computer, and actuator drive power supplies. A genetic algorithm is adopted as a control algorithm to obtain an optimum surface profile of the deformable mirror. A circular cross-sectional Gaussian beam was transformed into a beam with a rectangular contour using this system. Although the transfer function of this system is complicated and unknown, this system can be used to obtain the optimum beam profile within the achievable limits of a deformable mirror.     

Dynamic optical limiting experiments on vanadium dioxide and vanadium pentoxide thin films irradiated by a laser beam

Vanadium dioxide (VO2) and vanadium pentoxide (V2O5) thin films are irradiated by a near-infrared continuous-wave laser beam and the dynamic optical limiting performance is measured. The temperature varying with time of the films induced by a laser beam is also recorded by an IR thermal sensor. Under the irradiation of a laser beam with an intensity of 255 W/cm2 and a spot diameter of 2 mm, the laser beam transmittance of the VO2 film decreases from 47% before phase transition to 28% after phase transition, and the response time is approximately 200 ms; the laser beam transmittance of the V2O5 film decreases from 51% before phase transition to 24% after phase transition, and the response time is approximately 40 ms. The optical limiting is realized by this laser heating-induced phase transition.