粒子群和GMM-SDR模型在重叠谱峰解析中的应用

针对X光谱分析研究中经常遇到的谱峰重叠问题,提出了一种基于标准差关联的高斯混合模型(GMM-SDR)和粒子群算法的重叠谱峰解析方法。首先,介绍了重叠峰的GMM-SDR模型,并以GMM-SDR参数构成粒子位置,给出了粒子目标函数及适应度值的快速算法;然后,利用粒子群算法的群体搜索能力,以搜索最优GMM-SDR模型,进而实现重叠峰的分解。初始值采用随机设定,将测量的所有随机数据作为一个整体,以其对模型的概率匹配程度作为适应度值,故该方法避免了初值设定不当带来的局部收敛问题,克服了传统曲线拟合方法对原始有用数据的破坏,所搜索到的模型是一个全局最优解。通过对四个及以下重叠谱峰的分解表明,该方法分解精度较高,其中,两峰重叠谱分解后的峰位、峰面积及标准差最大相对误差分别为0.4%,0.05%和2.07%,三峰重叠谱分解后的峰位、峰面积及标准差最大相对误差分别为1.2%,0.04%和0.74%,可适用于各种严重重叠谱峰的分解。     

用广角X射线衍射系列方法测试和表征高聚物各种相态的结构特征

系统地论述了广角X射线衍射系列方法 :径向分布函数、圆柱分布函数、取向分布函数、Rietveld粉末衍射全谱图最小二乘拟合、纤维全谱图精修以及衍射全谱图分峰等方法的基础理论、实验方法、计算程序及其可获得的结构参数 ;阐明采用这些方法研究高聚物非晶相、取向非晶相、结晶相的进展情况 ;探讨了联用这些方法研究高聚物二相、三相共存结构的一种新思路和新方案 ,而且已取得一定进展 ;并展望这一研究框架的全面实现 ,将能进一步拓展高聚物共混物结构的研究     

太赫兹波段皮肤组织德拜参数提取算法的比较

针对非线性最小二乘法的传统算法在皮肤组织太赫兹波段德拜模型参数提取时,对参数初始值的要求很高,且所得模型与原始数据偏差较大的问题。将遗传算法应用于非线性最小二乘法对原始数据进行拟合,并提取参数。遗传算法随机产生初始解,不需要设置参数初始值,能简化参数提取过程;其搜索过程是从多个可能解开始,而不是单一初始值,所以不易陷入局部最优,基本能实现全局优化。以皮肤组织德拜参数提取为例,将遗传算法应用于非线性最小二乘法进行参数提取时不仅减小了文献中的模型在0.15~0.6THz频段与原始数据的偏差,而且在整个频率范围与原始数据都能较好吻合,验证了遗传算法提取德拜参数的有效性和皮肤组织太赫兹波段德拜模型的正确性。     

用广角X射线衍射系列方法测试和表征高聚物各种相态的结构特征

系统地论述了广角X射线衍射系列方法：径向分布函数、圆柱分布函数、取向分布函数、Rietveld粉末衍射全谱图最小二乘拟合、纤维全谱图精修以及衍射全谱图分峰等方法的基础理论、实验方法、计算程序及其可获得的结构参数;阐明采用这些方法研究高聚物非晶相、取向非晶相、结晶相的进展情况;探讨了联用这些方法研究高聚物二相、三相共存结构的一种新思路和新方案。而且已取得一定进展;并展望这一研究框架的全面实现,将能进一步拓展亢矣物共混物结构的研究。     

最优化算法在EDXRF谱线拟合中的应用

在X射线荧光光谱(EDXRF)测试中,由于探测器分辨率的影响,谱线会有不同程度的展宽,多元素存在时谱线还会有一定的重叠。获得纯净的谱线强度,并保证准确测试结果的过程,通常被称为EDXRF解谱。传统的解谱方法包括感兴趣区加和、纯元素谱线剥离以及干扰系数法等,但这些方法在解决EDXRF解谱问题时都有不同程度的局限性。最优化算法本质上是一种多自变量求极值的方法,首先列出具有多个自变量参数并能够描述真实物理学过程的目标方程,然后设定参数初值和边界条件,通过数学(如共轭梯度法)的运算,得到最优化的目标方程解。将最优化计算的思想应用于EDXRF解谱过程中,假设EDXRF谱图是一系列理想高斯峰的叠加,可以列出其与原始谱图残差的目标方程,根据物理学现象对这些高斯峰的高度、位置和宽度三个参数分别进行估计,此时目标方程的值通常较大,通过设定所有参数各自的边界条件,用共轭梯度法不断调整,做最优化计算,直至该值达到极小,此时多峰叠加的结果与实测谱最为贴合,解谱精度大大提高。利用这种方法研究开发了能用于商业化EDXRF仪器的软件程序。对最优化计算用于EDXRF解谱的方法进行了介绍,并以Pr和Nd混合液的L系EDXRF谱图的解析举例,这段谱线由11个谱峰组成,通过设定33个高斯峰参数的初始值,用共轭梯度法执行33个自变量的最优化计算,运行于普通计算机,经过580 ms的计算,拟合谱与实测谱的残差从37.645减小为1.6994,二者在对比谱图上也极为吻合,从而说明这是一种比较有效的谱线拟合方法。通过对结果的分析还发现,部分谱线的宽度发生了变化,真实地反映了该条谱线是由多条相邻谱线构成的事实。研究创新性在于,将数学中的最优化计算原理应用于EDXRF谱图解析过程中,获得了较好的效果,并以较为复杂的Pr和Nd双稀土元素的L系谱线解析为例进行了说明。     

热丝CVD生长金刚石薄膜喇曼散射研究

本文报道热丝化学气相沉积法（ＨＦＣＶＤ）生长金钢石薄膜的喇曼散射结果。选取多种峰型，对金钢石薄膜喇曼谱（１１０－１８００ｃｍ－１）采用最小二乘法进行非线性拟合，得到最佳拟合模型，其计算得到的拟合曲线怀实验谱图符合得较好。该模型揭示，石墨Ｄ峰（１３５５ｃｍ－１）是金刚石薄膜喇曼谱中不可缺少的一个组份，并且结合石墨Ｄ峰和金刚石喇曼的空间相关线型，可以解释金刚石喇曼区特殊峰形的物理机制，拟合参量的进一步     

阻尼最小二乘法用于激光诱导击穿光谱重叠特征谱线分离提取

近年来,激光诱导击穿广谱技术发展迅速。作为一种用于物质成分探测的新技术,它具有简单、快速、无需复杂样品制备、多种元素同时检测等优点,可实现待测样品物质成分现场、在线的检测,在很多领域都极具应用前景。激光诱导击穿光谱特征谱线的分离拟合提取是光谱特征识别与后续元素浓度定量反演研究的基础。为实现激光诱导击穿光谱重叠特征谱线的有效分离拟合提取,采用阻尼最小二乘法,分析并确定了迭代前的初始拟合参数值,实现了在重叠特征谱线情况下对某火力发电厂粉煤灰中的铬元素特征谱线的分离提取。阻尼最小二乘法基于高斯-牛顿迭代,在迭代步长中引入阻尼因子,在迭代的过程中根据每一步迭代后所反馈的信息动态的调整迭代步长,从而有效防止了迭代的发散,保证了迭代的快速收敛,最终使得元素特征谱线拟合提取的效果更佳、所得到的特征谱线强度值更准确。分别采用阻尼最小二乘法和最小二乘法对不同浓度的样品中铬元素特征谱线进行分离拟合提取并给出特征谱线的强度值,作出特征谱线强度值关于元素浓度的定标曲线并对比两种方法所得结果的线性相关性。结果表明,阻尼最小二乘法所得结果的线性相关性更高,该方法稳定、可靠,适用于激光诱导击穿光谱重叠特征谱线的分离拟合提取。     

基于连续小波变换的FTIR光谱拟合算法

提出了一种基于连续小波变换的FTIR光谱拟合算法。在计算差减尺度因子时,同时考虑原始谱图及其连续小波变换谱图,从光谱最小二乘拟合的角度求解,克服了常规差谱算法中的参考峰及差减因子的人工选择问题。采用六种不同的小波进行光谱拟合,用计算得到的差减因子来定量酒精度,误差绝对值的平均值仅为0.047°~0.072°,误差标准差仅为0.056°~0.091°。实验结果表明,连续小波变换结合最小二乘拟合的光谱拟合模型能为FTIR差谱提供一种准确可靠的新方法。     

基于BOTDA布里渊背向散射光谱数据的拟合算法

BOTDA布里渊背向散射光谱数据表现为洛伦兹型曲线,对不同的单模光纤和多模光纤会生成不同的波峰个数,单模光纤中的布里渊背向散射光谱表现为单个波峰,采用非线性最小二乘法对单峰进行拟合,采用高斯-牛顿迭代法求最小二乘解。迭代运算的效率取决于初始值的精度,当误差较大时,迭代次数会增加,甚至无法收敛。采用传统法和定积分法分别获取某一初始值。实验结果表明,在满足同等精度的条件下,定积分法能降低迭代次数,对布里渊背向散射光谱的单峰拟合具有较好的效果。     

阻尼最小二乘算法CARS光谱温度拟合

为了解决相干反斯托克斯喇曼散射（CARS）光谱CARSFT计算程序在温度拟合过程中,对初值依赖大、测量光谱信噪比差、不易收敛的问题,通过分析非线性迭代算法,采用了阻尼最小二乘算法对氮气CARS光谱进行温度拟合。针对CARS理论光谱计算公式较为复杂的特点,采用数值差分的方式来获得迭代计算时所需要的设计矩阵。分析了横向偏移、纵向偏移和温度三个主要参量对拟合残差的影响,发现纵向偏移的初值设置对温度拟合影响较大。拟合温度2000 K时的标准理论CARS光谱,设置偏离较大的初始值,采用阻尼最小二乘法获得了较好的结果。迭代计算167步后,温度拟合为2 005.6 K,残差为0.027 463。拟合光谱信噪比较差的CARS光谱,阻尼最小二乘法也能稳定收敛。结果验证了阻尼最小二乘法对初值的依赖不大,并且当拟合谱信噪比较差时也能收敛,可用于在恶劣环境下CARS测量光谱的拟合。     

基于模拟退火的相位展开方法

二维相位展开技术被广泛应用于光学精密测量、自适应光学、合成孔径雷达、图像处理等领域中。提出了一种基于模拟退火的相位展开方法。此方法对主值相位图采用局部平面近似，用模拟退火算法求出最优化平面参量，从而得到去除噪声和2π相位跳变的真实相位图。计算机模拟和实验结果均证明，此方法能有效地去除相位图中的噪声，实现带噪声的主值相位图的相位展开，可靠地重建被测物体的表面形貌。详细介绍了这一方法的原理及实验结果。     

Multivariate empirical mode decomposition approach for adaptive denoising of fringe patterns

An adaptive approach is presented for noise reduction of optical fringe patterns using multivariate empirical mode decomposition. Adjacent rows and columns of patterns are treated as multichannel signals and are decomposed into multiscale components. Fringe patterns are reconstructed with less noise by simply thresholding coefficients in different scales. The proposed approach can better concentrate local main components of fringe signals into single scale, compared with the conventional multiscale denoising method. A simulated pattern and an actual example are examined. Signal-to-noise ratio (SNR) of the simulated pattern is more than doubled.     

A New Chaotic Parameters Disturbance Annealing Neural Network for Solving Global Optimization Problems

Since there were few chaotic neural networks applicable to the global optimization, in this paper, we proposea new neural network model - chaotic parameters disturbance annealing (CPDA) network, which is superior to otherexisting neural networks, genetic algorithms, and simulated annealing algorithms in global optimization. In the presentCPDA network, we add some chaotic parameters in the energy function, which make the Hopfield neural network escapefrom the attraction of a local minimal solution and with the parameter p1 annealing, our model will converge to theglobal optimal solutions quickly and steadily. The converge ability and other characters are also analyzed in this paper.The benchmark examples show the present CPDA neuralnetwork's merits in nonlinear global optimization.     

Using hysteresis for optimization

We propose a new optimization method based on a demagnetization procedure well known in magnetism. We show how this procedure can be applied as a general tool to search for optimal solutions in any system where the configuration space is endowed with a suitable "distance." We test the new algorithm on frustrated magnetic models and the traveling salesman problem. We find that the new method successfully competes with similar basic algorithms such as simulated annealing.     

Quantum annealing by ferromagnetic interaction with the mean-field scheme

Quantum annealing is a novel method for combinatorial optimization problems. In this paper, we discuss the appropriate choice of quantum fluctuations in quantum annealing. The existence of room of choices of quantum fluctuations is an advantage of quantum annealing over simulated annealing. We consider the ferromagnetic interaction as a source of quantum fluctuations. Using the mean-field annealing scheme, we show that quantum annealing by ferromagnetic interaction is more efficient than the conventional quantum annealing and simulated annealing in the ground state search of the random-field Ising model.     

Deducing 1D concentration profiles from EPR imaging: a new approach based on the concept of virtual components and optimization with the genetic algorithm

Application of the genetic algorithm (GA) in conjunction with the concept of virtual components (VC) to determine 1D concentration profiles from EPRI spectra (images) is described. In this approach the concentration profile is expressed as the superposition of virtual components described by analytical functions of the Gaussian and Boltzmann type. The method was implemented in the computer program ACon, which allows for fully automated profile extraction via the nonlinear least-squares fitting of experimental images. The parametric sensitivity of the GA internal parameters such as population size, probabilities of the crossover, mutation and elitist retention to the search space was investigated in detail in order to find their optimal settings. The customized genetic algorithm was evaluated using simulated and experimental test data sets and its performance was compared with the Monte Carlo approach.     

超结构换热器网络综合优化的遗传/模拟退火算法-Ⅱ.遗传/模拟退火算法及应用

本文提出了用于换热器网络综合的改进的遗传/模拟退火算法.最优综合的模型基于本文第一部分所提出的通用解方法.所采用的遗传算法结合了模拟退火算法和爬山优化算法,同时引进精英策略和结构变异策略以增强算法的搜索能力.采用本文提出的算法对文献提出的算例进行了计算并得到了更好的结果.     

大尺度分层介质电特性参数的反演方法研究

针对大尺度分层介质电磁场散射回波数据, 提出了一种用于分层介质特性参数反演的新方法, 实现了分层介质的介电常数、电导率和介质厚度的同时反演. 方法首先将这些参数的反演转化为最优化问题进行求解, 然后应用模拟退火算法搜索最优解, 充分利用算法的全局寻优能力, 同时改善了算法的搜索控制策略, 使算法在最优解的搜索过程中可以自适应地调整搜索步长, 提高了算法的搜索效率. 结果表明, 本文方法可以实现大尺度分层介质特性参数的准确反演, 且具有较强的抗噪声能力. 该方法可以应用于火星/月球 雷达探测回波数据的反演以及地下分层结构特性的分析. 关键词： 电磁散射 分层介质 模拟退火算法 最优化     

Global and local contrast enhancement algorithm for image using wavelet neural network and stationary wavelet transform

A new contrast enhancement algorithm for image is proposed employing wavelet neural network (WNN)and stationary wavelet transform (SWT). Incomplete Beta transform (IBT) is used to enhance the global contrast for image. In order to avoid the expensive time for traditional contrast enhancement algorithms,which search optimal gray transform parameters in the whole gray transform parameter space, a new criterion is proposed with gray level histogram. Contrast type for original image is determined employing the new criterion. Gray transform parameter space is given respectively according to different contrast types,which shrinks the parameter space greatly. Nonlinear transform parameters are searched by simulated annealing algorithm (SA) so as to obtain optimal gray transform parameters. Thus the searching direction and selection of initial values of simulated annealing is guided by the new parameter space. In order to calculate IBT in the whole image, a kind of WNN is proposed to approximate the IBT. Having enhanced the global contrast to input image, discrete SWT is done to the image which has been processed by previous global enhancement method, local contrast enhancement is implemented by a kind of nonlinear operator in the high frequency sub-band images of each decomposition level respectively. Experimental results show that the new algorithm is able to adaptively enhance the global contrast for the original image while it also extrudes the detail of the targets in the original image well. The computation complexity for the new algorithm is O(MN) log(MN), where M and N are width and height of the original image, respectively.     

Ranking important nodes in complex networks by simulated annealing

In this paper, based on simulated annealing a new method to rank important nodes in complex networks is presented.First, the concept of an importance sequence(IS) to describe the relative importance of nodes in complex networks is defined. Then, a measure used to evaluate the reasonability of an IS is designed. By comparing an IS and the measure of its reasonability to a state of complex networks and the energy of the state, respectively, the method finds the ground state of complex networks by simulated annealing. In other words, the method can construct a most reasonable IS. The results of experiments on real and artificial networks show that this ranking method not only is effective but also can be applied to different kinds of complex networks.