高能激光准直传输远场斯特雷尔比与数值验证

利用扩展惠更斯-菲涅耳原理对高能激光在大气中的传输进行了分析,获得了高能激光准直传输轴上及峰值斯特雷尔比的积分表达式,并使用激光大气传输四维仿真程序对该表达式的适用范围进行了分析及数值模拟验证。结果表明,积分表达式可以很好地适用于无自适应光学系统补偿时的高能激光湍流热晕传输问题,可以用于自适应光学系统补偿时弱热晕条件下的湍流热晕角非等晕问题和弱热晕及中等热晕条件下的湍流热晕聚焦非等晕问题。     

热晕瞬态效应的数值模拟

根据瞬态四维（４Ｄ）热晕程序，研究了激光大气传输中热晕瞬变效应及如何过渡到等压近似下的热晕状态，并与等压近似下４Ｄ热晕程序计算结果作了对比，在大约四倍流体力学时间，瞬态４Ｄ热晕程序计算结果逼近等压近似下的结果；另外，还作了相位补偿计算，对光强每平方厘米几千瓦的情况，在理想的自适应光学系统动力学条件下，瞬变阶段的热畸变能相当有效地实时补偿。     

再生反馈技术在跟踪控制系统中的应用

为提高伺服系统跟踪精度，利用现代计算机辅助控制技术，用软件进行坐标变换与反变换、数字计算与滤波和再生补偿量的计算，构成具有再生反馈(近似前馈)的复合控制跟踪系统，最终实现对高速运动目标的高精度跟踪。此技术应用于国家某重点项目中，经过陆地试验和海上试验，充分验证了试验效果很理想。本技术可应用于各种运动目标的跟踪系统中，可有效提高系统跟踪精度。     

非线性热晕效应自适应光学位相补偿

采用已建立的一套较完整的包括真实自适应光学位相补偿系统仿真的高能激光大气传输的数值模拟程序，对非线性热晕应相偿进行了计算。分析了非线性热晕效应位相补偿不稳定性的基本特征。     

结合人工神经网络的翅片管式冷凝器快速仿真模型

为了提高翅片管式冷凝器仿真精度和计算速度，提出了含基本数学模型与两个神经网络的复合冷凝器模型．基本数学模型采用分区集中参数模型，简化用神经网络用以反映分区集中参数模型与分布参数模型间的非线性关系，精度改进用神经网络通过分析比较模型与实验值的差别，提高复合模型的精度．实际应用表明，复合模型的计算速度比分布参数     

稳态热晕非等晕效应的数值分析

运用高斯光束展开的方法,分析圆对称平顶光束在大气传输中的热晕及其非等晕效应。通过对热晕及非等晕效应引起的波前畸变作泽尼克多项式展开,得到了各阶泽尼克系数,进而求得热晕校正残差,数值计算得到的热晕拟合校正残差与现有理论公式得到的结果基本一致。对非等晕效应的数值模拟结果表明:在热畸变参数较小,且给定口径尺寸的情况下,热晕角度非等晕校正残差与非等晕角θ的平方成正比,热晕聚焦非等晕校正残差与信标高度的1.71次幂成反比。     

非共轴光学系统传递函数计算

用坐标变换计算非共轴光学系统是十分有效的。本文讨论了非共轴光学系统传递函数(OTF)的计算要点,并给出了平行光束传递函数的计算方法。计算了一个无焦系统的波动光学传递函数。     

全反射棱镜式激光陀螺自适应稳频技术

针对全反射棱镜式激光陀螺在稳频过程中相敏信号易受噪声干扰的现象,研究了全反射棱镜式激光陀螺自适应稳频技术。理论分析了全反射棱镜式激光陀螺的稳频特性,结合自适应噪声对消原理,建立了全反射棱镜式激光陀螺自适应稳频系统的数学模型。通过硬件电路设计,搭建了基于递归最小二乘算法的自适应稳频控制系统。分别对原有稳频技术与自适应稳频技术进行了实验测试,实验结果表明,自适应稳频能有效消除噪声对相敏信号的干扰,稳频精度提高了近一个数量级,陀螺精度相应提高了60%以上。此分析结果为提高全反射棱镜式激光陀螺的性能提供了重要参考。     

热晕效应模式法自适应光学校正的数值模拟

改进了泽尼克模式波前重构数值模拟方法,避免了传统方法中大量的面积分运算,对准直光束瞬态热晕效应模式法自适应光学校正进行了数值模拟研究。数值模拟结果表明,由于模式混淆误差的影响,使泽尼克重构模式阶数的选择受限,即存在一个最大的重构模式阶数,大于该模式阶数会导致校正效果迅速变差,且最大模式阶数的大小与热晕强度无关。对于子孔径分布为8×8的69单元自适应光学系统,最大模式阶数为37阶;同直接斜率法相比,泽尼克模式法可以通过重构模式选择,提高自适应光学系统对热晕校正的稳定性,抑制热晕补偿不稳定性的发生。     

全反射棱镜式激光陀螺自适应稳频技术EI北大核心CSCD

针对全反射棱镜式激光陀螺在稳频过程中相敏信号易受噪声干扰的现象,研究了全反射棱镜式激光陀螺自适应稳频技术。理论分析了全反射棱镜式激光陀螺的稳频特性,结合自适应噪声对消原理,建立了全反射棱镜式激光陀螺自适应稳频系统的数学模型。通过硬件电路设计,搭建了基于递归最小二乘算法的自适应稳频控制系统。分别对原有稳频技术与自适应稳频技术进行了实验测试,实验结果表明,自适应稳频能有效消除噪声对相敏信号的干扰,稳频精度提高了近一个数量级,陀螺精度相应提高了60%以上。此分析结果为提高全反射棱镜式激光陀螺的性能提供了重要参考。     

Compressed Sensing Quantum State Tomography Assisted by Adaptive Design

The recently proposed compressed sensing(CS) method sheds light on quantum state tomography of multi-qubit systems with low rank, which greatly reduces the complexity of measurement and computation. However, the restricted isometry property requirement of CS is difficult to be promised or verified in practice, which makes this method probably assign unreasonable results. In regard to this problem, we adopt a two-step procedure and implement an adaptive strategy to update measurement operators based on the measurement results of the first step for CS, which not only serves as a way to verify the estimate but also improves the accuracy of tomography.Our numerical simulations manifest that our adaptive protocol can reduce about half of the infidelity of nonadaptive protocol and is still efficient even when the rank of the state is slightly high, which would greatly benefit multi-qubit state tomography in future experiments.     

二维泊松方程的遗传PSOR改进算法

针对二维泊松方程在实际应用过程中几种常用方法存在计算量大、易发散、局部收敛等不足,提出了一种改进算法.该算法基于并行超松弛迭代法,采用遗传算法对松弛因子进行全局寻优,解决了超松弛迭代法求解泊松方程时最佳松弛因子难以确定的问题.构建了多目标适应度函数,优化了遗传算子参数,分析了算法的计算量、计算时间与误差精度,与传统方法进行了对比研究.结果表明:松弛因子对泊松方程求解的速度与精度影响显著;改进算法能减少迭代次数,节省计算时间,加快方程的求解;算法适合于求解计算量较大、精度要求较高的时域有限差分方程,而且精度要求越高,算法的性能越好,节省的时间也越多.     

利用改进的符号数算法和光学符号代换实现矩阵计算

本文提出了一种利用改进的符号数算法和多窗口解码光学符号代换法则实现多值矩阵计算的光学方法。并给出两个多比特改进的符号数矩阵外积计算的实验结果。这一方法具有精度高、速度快等特点。     

Long-range interactions in Monte Carlo simulation of confined water

We investigate methods for the treatment of long-range interactions in the context of grand canonical Monte Carlo (GCMC) simulations of water adsorption in slit-shaped activated carbon pores. Several approaches, ranging from the simple minimum image convention to the more complex two-dimensional Ewald summations, are implemented and compared with respect to accuracy and speed of computation. The performance of some of these methods in GCMC is found to be significantly different from that in molecular dynamics applications. Of all the methods studied, one proposed by Heyes and van Swol was the most promising, providing the best balance between accuracy and speed. In our application, it was shown to be about 2 times faster than the fastest of the two-dimensional Ewald methods. We expect this conclusion to apply in general to systems that are periodic in two dimensions and finite in the third.     

An improved symplectic precise integration method for analysis of the rotating rigid–flexible coupled system

This paper presents an improved symplectic precise integration method (PIM) to increase the accuracy and keep the stability of the computation of the rotating rigid–flexible coupled system. Firstly, the generalized Hamilton's principle is used to establish a coupled model for the rotating system, which is discretized and transferred into Hamiltonian systems subsequently. Secondly, a suitable symplectic geometric algorithm is proposed to keep the computational stability of the rotating rigid–flexible coupled system. Thirdly, the idea of PIM is introduced into the symplectic geometric algorithm to establish a symplectic PIM, which combines the advantages of the accuracy of the PIM and the stability of the symplectic geometric algorithm. In some sense, the results obtained by this method are analytical solutions in computer for a long span of time, so the time-step can be enlarged to speed up the computation. Finally, three numerical examples show the stability of computation, the accuracy of solving stiff equations and the capability of solving nonlinear equations, respectively. All these examples prove the symplectic PIM is a promising method for the rotating rigid–flexible coupled systems.     

对正交浮雕光栅傅里叶模方法的快速实现

采用“逆规则”傅里叶展开法来提高正交浮雕双周期光栅问题的收敛性,把用在单周期光栅中的反射透射系数阵算法推广到正交双周期光栅中,并充分利用光栅的对称性和对角周期性,在保证计算准确度和稳定性的条件下,把计算速度提高了几十倍,从而把对这类问题计算从大型计算机中解脱出来,而可用微机加以处理。     

Laser heating of optically thin samples

Summary A mathematical model able to describe the temperature profiles generated in a thin film by the steady-state illumination by a Gaussian laser beam is presented. The film is supposed to be made by a weakly absorbing liquid sample bounded by two parallel transparent plates, the cell walls, whose thermal exchange to the surrounding ambient may be assumed to be linear with the temperature difference. An analytical solution of the problem is presented in form of Hankel transform and a matrix numerical approach to the computation of the temperature profiles is reported. It is particularly well suited to a computer implementation and allows one to get very accurate results in very short computing times. The influence of the heat exchange coefficient to the ambient is shown in an example. A check of the method accuracy is performed by comparison with literature results in the particular case of infinite heat exchange coefficient. To speed up publication, the authors of this paper have agreed to not receive the proofs for correction.     

A Quick Numerical Method for Analyzing Nonlinear Effects of Excited State Absorption

１　Ｉｎｔｒｏｄｕｃｔｉｏｎ　　Ｗｈｅｎｕｌｔｒａｓｈｏｒｔ,ｈｉｇｈｉｎｔｅｎｓｉｔｙｌａｓｅｒｐｕｌｓｅｓｉｎｔｅｒａｃｔｗｉｔｈａｍｕｌｔｉｅｎｅｒｇｙｌｅｖｅｌｍｏｌｅｃｕｌａｒｓｙｓｔｅｍ,ｔｈｅｂｕｉｌｄｕｐｏｆｐｏｐｕｌａｔｉｏｎｉｎｅｘｃｉｔｅｄｓｔａｔｅｓｗｉｌｌｌｅａｄｓｔｏａｒｅｄｕｃｔｉｏｎｏｒａｎｉｎｃｒｅａｓｅｏｆｔｈｅａｂｓｏｒｐｔｉｏｎｃｏｅｆｆｉｃｉｅｎｔ,ｗｈｉｃｈａｒｅｒｅｆｅｒｒｅｄｔｏａｓｓａｔｕｒａｂｌｅａｂｓｏｒｐｔｉｏｎ（ＳＡ,ｉ．ｅ．,ａｎ…     

求解变量与离散体元界面速度选择对SIMPLE算法收敛特性的影响分析

针对复杂区域中流动计算的SIMPLE算法,在选用不同的速度分量作为求解变量和离散体元的界面速度时,进行了算法收敛速度及收敛准确性的对比分析。通过研究,澄清了现有算法中存在的概念失误,提出改进算法的措施,并给出算例。     

GPS“射线打靶”模式高效数值方法的研究及其并行实现

全球定位系统(GPS)"射线打靶法"是GPS MET(气象)资料变分同化中一种自成一体的观测算子,由于其计算量非常巨大而一直没能在资料同化中得到使用.为了减少该观测算子的计算量,针对GPS射线轨迹方程为一可分Hamilton系统的特点,采用二阶经济辛格式求解,与经典的四阶Runge Kuntta法相比,不仅计算准确度没有降低甚至有所改善,而且节省近4倍的计算时间.同时进一步研究了改进后的GPS"射线打靶"模式的并行实现.