光学薄膜的椭圆偏振模型分析与数据处理

通过对光学薄膜的椭偏光谱分析,研究了光学薄膜系统的模型（结构模型和色散模型）建立方法,并应用模拟退火算法（SA）来反演椭偏光谱测量数据.介绍了程序算法的建立与程序的编制,并对标准薄膜样品和自制样品进行测量和数据处理.实验结果表明,在可见光波段,折射率误差小于0.01,消光系数误差小于3×10-4.     

粒子群-模拟退火混合算法在群控电梯中的应用

针对高层建筑电梯多、分布散、维修不及时等问题,提出了一个优化调度、报警维修及时的群控电梯系统;采用多目标优化方法建立数学模型,利用粒子群算法的概念简单、收敛速度快、易于实现的优点,同时引入模拟退火思想来克服粒子群算法易陷入局部最优的缺陷,对目标函数进行优化,PLC编程计算出参数,上位机调度计算,根据算法结果,优化电梯,合理调度;经过MATLAB仿真分析表明,该系统节省了平均候梯时间、平均乘梯时间和系统能耗,缩短了故障时间和维修时间,具有较大的应用前景。     

薄膜光学常数的粒子群算法

为解决椭偏法测量薄膜厚度和折射率实验数据处理较为复杂的问题,采用一种新的基于群体智能的优化算法——粒子群算法处理实验数据.以单层吸收薄膜的测量为例,利用该算法进行数据处理.实验结果表明,可以同时获得3个薄膜参数(折射率n,消光系数k和薄膜厚度d),而且在确切参数范围未知情况下,大范围内进行搜索仍然能保证快速收敛到最优解.该算法与遗传算法以及利用椭偏仪数据处理软件得出的结果相比较,计算精度高,收敛速度快.     

基于粒子群优化算法的膜系设计方法

粒子群优化算法是一种新的演化计算技术,与遗传算法相比,粒子群优化算法具有易于实现,控制参量少等优点,且在大多数的情况下,可快速收敛于最优解.为了获得更优的膜系结构,本文提出了一种利用粒子群优化算法进行膜系设计的方法,并以增透膜、高反膜及分光膜为优化设计实例验证该方法的可行性.在这些实例中,以理论反射率和实际反射率的误差平方和为评价函数.结果表明,将粒子群优化算法用于膜系设计是有效的,在相同设计条件下,应用粒子群优化算法可以得到比遗传算法更优的膜系结构.     

基于粒子群优化算法的膜系设计方法

粒子群优化算法是一种新的演化计算技术,与遗传算法相比,粒子群优化算法具有易于实现,控制参量少等优点,且在大多数的情况下,可快速收敛于最优解.为了获得更优的膜系结构,本文提出了一种利用粒子群优化算法进行膜系设计的方法,并以增透膜、高反膜及分光膜为优化设计实例验证该方法的可行性.在这些实例中,以理论反射率和实际反射率的误差平方和为评价函数.结果表明,将粒子群优化算法用于膜系设计是有效的,在相同设计条件下,应用粒子群优化算法可以得到比遗传算法更优的膜系结构.     

基于多算法优化SVM的粗糙面参数反演

支持向量机（SVM）是粗糙面参数反演中常用的一种反演算法,SVM反演中的惩罚参数C和核函数参数G对反演结果精度的影响较大,若参数取值不当,会使模型产生\"过学习\"或者\"欠学习\"的现象,从而降低预测精度.给出几种SVM参数C和参数G的优化算法,如K折交叉验证（K-CV）、遗传算法（GA）和粒子群算法（PSO）,并在此基础上提出一种基于K-CV和GA改进的PSO算法（GA-CV-PSO）.利用矩量法（MoM）获得的粗糙面后向散射系数构造训练集和测试集,通过不同参数反演的仿真结果对比不同优化算法的反演精度和计算时间,表明GA-CV-PSO算法克服了单一优化算法的缺陷,具有更精确的反演精度和更强的泛化能力.     

宽波段波状多层膜结构偏振分束器的设计与优化

本文基于二维光子晶体波状多层膜结构,提出了叠加两种不同膜厚的周期膜堆来拓宽偏振分束器有效带宽的方法.采用粒子群优化算法,建立偏振分束特性的评价函数,优化结构和薄膜厚度等参数,获得了中心波长565 nm,带宽220 nm,平均消光比大于30 dB的宽波段偏振分束器.采用时域有限差分方法分析了膜层顶角的角度敏感性和波状结构的电场分布.结果表明,两个周期膜堆组合的结构解决了禁带不连续的情况,而粒子群优化算法的使用加快了结果收敛,有效地扩展了偏振分束带宽.     

基于动态粒子群优化的目标跟踪算法

目标跟踪问题的关键在于如何寻找与目标运动状态匹配的运动模型。交互式多模型算法的模型集是根据先验信息确定的,它不随时间变化而变化,并且要求在模型集中任意时刻都存在描述目标运动模型。在实际中需要大量模型来描述运动。将粒子群优化和变结构多模型算法相结合,不仅能充分利用系统的实时量测信息,还能根据其先验信息调节优化算法结构。仿真表明,运用动态自适应粒子群优化算法实现模型集自适应,可以提高目标跟踪的精度和实时性。     

基于PSO-DE算法的污水处理优化控制研究

针对污水处理系统能耗过大,变量多,非线性和滞后严重等特点造成的控制困难问题,提出了基于改进型粒子群的算法的优化控制。粒子群算法具有自适应控制,全局搜索等优点但本身存在早熟收敛及在进化后期收敛速度慢等缺点,通过优势互补思想引入差分进化算法,新算法结合两者优势有效提高粒子在全局的寻优效率,建立对应的混合算法优化模型,并与普通粒子群算法优化进行比较,结果证明了该算法在保证出水水质的前提下做到降低能耗。     

纳米厚度薄膜外差椭偏测量技术的研究

结合激光外差干涉法和透射式椭偏测量原理,研究了一种快速、高精度测量纳米厚度薄膜光学参数的方法。计算并分析了复灵敏度因子随薄膜参数和入射角度的变化规律、椭偏参数的选择及容许测量误差。两个声光调制器产生20kHz的拍频,采用简单的直接比相方法即可获得优于0.1°的相位分辨率,而且测量系统中没有使用任何波片和运动部件,抗干扰能力强且测量过程完全自动化,适用于工业现场在线连续测量。实验数据和理论分析表明,此方法可以达到亚纳米级测量精度。     

Determination of optical constants of a biaxially anisotropic film by standard multiangle monochromatic ellipsometry

Possibilities of standard multiangle monochromatic ellipsometry in the determination of parameters of a uniform biaxially anisotropic layer are studied for the case of an arbitrary orientation of the two principal axes in the plane of incidence of the light beam and perpendicularity of the third axis to the plane of incidence. Using numerical simulation, it has been found that the measurement accuracy that is necessary in the determination of all the three principal components of the dielectric permittivity tensor ε and tilt angle of the axes using only angular dependences of ellipsometric parameters must be no worse than 0.0001°, which is far beyond the accuracy limits provided by present-day ellipsometers. If the tilt angle is known, standard multiangle monochromatic ellipsometry provides the determination of thickness and all three principal components of the dielectric permittivity tensor. This method allows one to determine the layer thickness and tensor component for the axis perpendicular to the plane of incidence, as well as the average value of components for the axes lying in the plane of incidence without involving the data about the tilt angle of the axes. This is demonstrated by an example of experimental data for biaxially anisotropic SiO _x films obtained by oblique deposition of silicon monoxide SiO evaporated in vacuum.     

Polarization modulation imaging ellipsometry for thin film thickness measurement

A polarization modulation (PM) imaging ellipsometer is proposed and setup in order to measure precisely the thickness of thin film. Five images are collected sequentially by CCD camera with respect to five pre-determined azimuth angles of a quarter wave plate (QWP) during measurement. Then two-dimensional (2-D) distributions of the ellipsometric parameters ψ and Δ over the full dynamic range are obtained. Conceptually, PM imaging ellipsometer integrates the features of phase shift interferometry with conventional photometric ellipsometry by rotating the QWP sequentially to produce polarization modulation that is able to measure the thickness of a thin film in two dimensions precisely and quickly. The basic principle of PM imaging ellipsometer is derived wherein features such as common path configuration, full dynamic range of measurement, and insensitive to non-uniform response of the CCD are analyzed. The experimental results verify the ability and performance of PM imaging ellipsometer on 2-D thin film thickness, while the errors regarding the ellipsometric parameters measurements are discussed.     

Copper phthalocyanine film studied with spectroscopic ellipsometry

The ellipsometric spectra of vacuum sublimed copper phthalocyanine (CuPc) thin film on single-crystal silicon were studied on a new type of scanning ellipsometer with the analyser and polarizer rotate synchronously. The electronic structure of the spectra is discussed.     

Optimization of conducting polymer thin film optical and electrical waveguides

The fabrication and characterization of conducting polymer thin film optical and electrical waveguides prepared by vacuum deposited method is reported. The waveguiding parameters (refractive index, propagation loss, V-I and C-V characteristics) have been measured.     

Out-of-plane ellipsometry measurements of nanoparticles on surfaces for thin film coated wafer inspection

Measurements of the diameter and size distribution of nanoparticles on wafers are critical parameters in the semiconductor industry, essential to control transistor quality and increase production rate. A goniometric optical scatter instrument (GOSI) has been developed that employs polarized light scattering to make measurements of the diameter and size distribution of nanoparticles on bare and thin film coated wafers. This scatter instrument is capable of distinguishing various types of optical scattering characteristics, which correspond to the diameters of the nanoparticles and thin film thickness, on or near the surfaces using the Mueller matrix calculation in Bobbert and Vlieger (1986) [1]. The experimental results of the GOSI system show good agreement with theoretical predictions for nanoparticles of diameter 100, 200, and 300 nm on wafers coated with thin films of 2, 5, and 10 nm thickness. These results demonstrate that the polarization of light scattered by nanoparticles can be used to determine the size of particulate contaminants on bare and thin film coated silicon wafers.     

用模拟退火算法研究非晶硅薄膜的光学性质

在玻璃衬底上采用等离子体增强的化学气相沉积(PECVD)法制备了非晶硅薄膜(A-Si:H)。用紫外-可见-近红外分光光度计测出了其透射光谱。采用模拟退火算法研究了透射光谱,得出了薄膜的厚度、折射率和吸收系数随波长变化的关系式、光学带隙等光学常数,并对该方法的优缺点进行了讨论。     

Summary of conditions for precision in applications of ellipsometry to surface and film studies

Ellipsometry [1,2] is simple in instrumentation and application and is sensitive to fractional monatomic coverage under favourable conditions. This summary recapitulates the requirements for fruitful applications of this technique. In general it will be applied to study an interface situation or a reaction at an interface. It will therefore be necessary to obtain the optical parameters of the bulk phase and reaction products known or suspected to be involved at the interface when the nature and kinetics of the reactions occuring there are studied. These can then be used to compute curves based on various model postulates which can then be compared with observations until an acceptable interpretation can be places on the observations. The determination of optical parameters for a film-free interface of a bulk phase often then turns out to be simply a question of preparation and preservation of such a film-free interface.     

Convergence test for inversion of frequency-resolved optical gating spectrograms

We describe a new and simple method to aid in the analysis of retrieved pulses from inverted frequency-resolved optical gating (FROG) traces. The analysis can separate noise from distortion and shows that distortion is more deleterious to the retrieved pulse than is pure noise. The analysis relies on the fact that FROG traces can be constructed from a single outer product of two vectors, whereas distortion and noise require the sum of a series of outer products.     

Explicit solution for the optical properties of a uniaxial crystal in generalized ellipsometry

The complex ordinary (N_o) and extraordinary (N_e) refractive indices of an absorbing uniaxial crystal can be explicitly derived from the normalized diagonal (α) and off-diago-nal (β) elements of one oblique-incidence (at angle φ) reflection matrix measured by generalized ellipsometry on a crystal face that contains the optic axis at an oblique orientation (ω ≠ O or $\text{π}\text{2}$) with respect to the plane of incidence. At most four solution sets (N_o, N_e) are mathematically and physically consistent with one measurement set (α,β,φ,ω). In many cases, identification of the correct solution is feasible without additional measurements. If necessary repeated measurements at a different value of φ or ω will resolve the ambiguity; only one solution set remains invariant upon a change of φ or ω, the correct set. A single measurement of reflectance (e.g., for p-polarized light) may also be adequate. The analytic inversion method is used in an error analysis to determine optimum choices of angle of incidence (φ ? 50–70°) and crystal orientation (ω ? 45°) that lead to minimum percentage errors of N_o and N_e for several uniaxial crystals.