基于改进粒子群-GRNN模型的地铁隧道沉降预测

为了实现地铁隧道沉降的准确预测,针对传统方法和常用智能方法预测精度不高、适用性不强等问题,建立基于改进粒子群和广义回归神经网络的隧道沉降预测模型。模型引入随机变异因子以克服粒子群算法早熟收敛和后期搜索效率不高的缺陷。通过与GRNN、普通PSO-GRNN和PSO-BP模型进行对比,验证了改进算法的有效性和所建模型的优越性。以长沙地铁隧道为例进行沉降预测发现:预测值与实测值相差1.04 mm,相对误差为4.05%,预测精度高,满足工程需要。     

基于粒子群优化算法神经网络在开采沉陷中的应用

考虑BP网络存在收敛速度慢、局部极值等缺点,引入线性下降惯性权重粒子群优化（LWPSO）算法,建立基于线性下降惯性权重粒子群优化（LWPSO）算法的人工神经网络模型,在分析抚顺发电有限责任公司厂区地表下沉的实际观测资料的基础上,对厂区的任意点,任意时刻进沉陷预测研究。     

多项目交互组合选择模型的改进粒子群优化研究

为解决多项目之间发生交互作用时的项目选择和组合优化问题,在已有优化模型的基础上,提出改进的粒子群算法,利用种群间跳跃改进的方法增强了粒子的搜索性能,并通过聚类和干涉的方法,防止算法陷入局部最优,有效求解多目标优化模型,最后利用算例分析证明模型和算法的有效性。     

时滞离散智能系统的动态输出反馈镇定控制器综合的统一方法

提出标准神经网络模型（SNNM）来描述包含神经网络或T—S模糊模型的时滞（或非时滞）离散智能系统．SNNM由离散线性动力学系统和有界静态非线性算子连接而成．利用SNNM的全局渐近稳定性分析的结果，分别设计线性或非线性动态输出反馈控制器，使得SNNM的闭环系统稳定．控制方程可以表示为线性矩阵不等式（LMI）形式，便于利用各种凸优化算法求解以获得控制规律．大部分基于神经网络（或模糊模型）的时滞（或非时滞）离散智能系统都可以转化为SNNM，以便采用统一的方法来综合这些智能系统的控制器．SNNM的3个应用例子表明：SNNM不仅使得大多数基于神经网络（或模糊模型）的离散智能系统镇定控制器的综合简单易行，而且为其他类型的非线性系统的控制器综合提供新的思路．     

基于粒子群寻优机制的图像分割方法

利用粒子群优化（eso）算法全局寻优、快速收敛的特点,结合模糊C-均值（FCM）图像分割算法提出一种新算法,用PSO算法代替了FCM算法的基于梯度下降的迭代过程,使算法具有很强的全局搜索能力,很大程度上避免了FCM算法易陷入局部极小的缺陷;同时也降低了FCM算法对初始值的敏感度。实验结果表明,与FCM相比该算法聚类更准确,效率更高,具有较高的分割速度和良好的抑制噪声的能力。     

基于速度向量化策略的二进制粒子群算法

基于二项分布等概率统计方法对原始的离散二进制粒子群算法进行分析,发现速度最大值和惯性权重这两个参数对原算法的搜索性能影响较大,且原算法的速度更新方式会降低种群多样性,导致全局搜索性能不强。基于此,提出基于速度向量化策略的二进制粒子群算法,通过对粒子的速度进行向量化,较好地提高了算法全局搜索性能。通过实例仿真对比,证明了所提算法的可行性及其优势。     

激活函数可调的神经元模型及其有监督学习与应用

提出一种激活函数可调的新神经元模型（tunable act ivation function，简记为TAF）模型，并给出这类模型的一般形式，该模型用于多层前向神经网络MFNN时，其激活函数可借类似BP算法进行训练而求得，通过几个具体例子给出了对激活函数进行训练的算法，试验结果表明，采用TAF模型的多层前向神经网络的网络容量和性能，优于采用通常M－P模型的网络。     

粒子群优化算法在确定河流水质参数中的应用

将粒子群优化（PSO）算法应用于求解分析瞬时投放示踪剂情况下的一维河流水团示踪试验数据，以及确定河流水质参数的函数优化问题。分别就粒子数目和待估水质参数的初始取值范围对算法运算过程的影响进行了数值实验。结果表明：①PSO算法能够有效地应用于求解分析河流水质试验数据，确定水质参数的函数优化问题；②粒子数目的多少对迭代次数、运算时间和算法是否收敛有一定的影响，在粒子数目较大的情况下，可以保证运算过程收敛；③待估参数初始猜测值的选取范围对迭代次数也有一定的影响，选取范围越大，需要的迭代次数越多。最后，指出了需要进一步研究的问题。     

基于GSFLA的电力系统谐波估计算法

针对非线性动态负载引起的谐波难于检测的问题,提出了一种新的谐波估计算法。该算法借助蛙跳算法（SFLA）的全局搜索性对未知参数进行优化估计;引入高斯分布估计算法（GEDA）的思想,对蛙群中适应度好的蛙进行分布估计再生,提高收敛速度;结合进化代数改进蛙跳规则以改善局部搜索性能。实验仿真数据显示,与PSO算法相比,振幅平均估计精度提高了5．3％,相角平均估计精度提高了4．7°。研究表明,该算法（GSFLA）用于电力系统的谐波估计有更快的收敛速度和估计精度。     

考虑质量设计的合同组批生产问题研究

针对在制定组批方案中可能会产生的质量浪费现象,在考虑质量设计因素的基础上,建立以质量离散度为优化目标的组批模型,进而设计出考虑质量设计的合同组批算法。该算法改进了粒子群算法以实现聚类组批,首先松弛设备容量限制,以合同的质量离散度最小为目标,运用粒子群算法进行聚类计算,得到初步的组批方案;然后考虑设备容量约束,基于质量离散度最小原则,对违背设备容量范围的批次进行修复调整,得到最终的组批方案。以某大型国有钢铁集团的无缝钢管厂为例,基于生产实际数据设计了数据实验,以质量离散度为评价指标,将本文算法与基本粒子群算法进行对比,探讨了批次数量对问题求解的影响。实验结果表明,本文提出的模型和算法是可行和有效的。     

膜分离处理印染废水研究进展

印染废水具有水量大、色度高、成分复杂、环境污染严重等特点。膜分离技术处理印染废水具有选择性好、生产效率高和处理成本低等特点。基于对近年来的文献调研,综述了膜分离技术在印染废水处理中的研究进展情况,指出了膜分离法处理印染废水还存在的主要问题和未来发展方向,并对膜分离技术处理印染废水应用前景作了展望。     

Modelling and Trading the English and German Stock Markets with Novelty Optimization Techniques

The motivation for this paper was the introduction of novel short‐term models to trade the FTSE 100 and DAX 30 exchange‐traded funds (ETF) indices. There are major contributions in this paper which include the introduction of an input selection criterion when utilizing an expansive universe of inputs, a hybrid combination of partial swarm optimizer (PSO) with radial basis function (RBF) neural networks, the application of a PSO algorithm to a traditional autoregressive moving model (ARMA), the application of a PSO algorithm to a higher‐order neural network and, finally, the introduction of a multi‐objective algorithm to optimize statistical and trading performance when trading an index. All the machine learning‐based methodologies and the conventional models are adapted and optimized to model the index. A PSO algorithm is used to optimize the weights in a traditional RBF neural network, in a higher‐order neural network (HONN) and the AR and MA terms of an ARMA model. In terms of checking the statistical and empirical accuracy of the novel models, we benchmark them with a traditional HONN, with an ARMA, with a moving average convergence/divergence model (MACD) and with a naïve strategy. More specifically, the trading and statistical performance of all models is investigated in a forecast simulation of the FTSE 100 and DAX 30 ETF time series over the period January 2004 to December 2015 using the last 3 years for out‐of‐sample testing. Finally, the empirical and statistical results indicate that the PSO‐RBF model outperforms all other examined models in terms of trading accuracy and profitability, even with mixed inputs and with only autoregressive inputs. Copyright © 2016 John Wiley & Sons, Ltd.     

基于支持向量机和粒子群算法的电子鼻伤口感染检测

针对传统的伤口感染诊断方法耗时长,操作复杂等问题,提出了一种基于电子鼻和支持向量机(SVM)的方法进行伤口感染检测,分别检测非感染和三种常见病原菌感染的大白鼠伤口顶空气体,然后利用 SVM对实验数据进行识别.同时,鉴于传感器阵列的优化以及 SVM参数选择对其分类准确率有重大的影响,提出一种基于粒子群算法(PSO)的传感器阵列和 SVM参数同步优化方法.实验结果表明,SVM结合 PSO与传统的神经网络以及遗传算法相比,极大提高伤口感染检测的准确率     

嫦娥三号巡视器视觉定位方法

月面巡视器的定位是巡视器开展月面科学考察工作的基础,是一项关键技术.本文提出了一种基于计算机视觉的定位方法,将SIFT(scale-invariant feature transform)匹配、相关系数匹配、最小二乘匹配和光束法平差等多项技术融合,实现了相邻站间月面巡视器的导航定位.在实验室构建立体视觉导航系统对本文的方法进行可行性和精度测试,结果表明视觉定位相对精度将优于4%,并成功应用于玉兔号巡视器定位.     

有色噪声下随机时滞传染病模型的稳定性

本文运用O-U过程刻画环境变化性,在Edoardo Beretta基础上构造了有色噪声影响下的随机时滞的传染病模型。运用一般Lyapunonv方法研究了有色噪声对该系统的影响并得到系统正平衡点保持稳定的充分条件。最后通过对比发现随机扰动对系统稳定性影响仅仅与其随机过程的方差有关。     

递推式视频前景分割方法

本文描述了一种基于前一帧已有分割结果对当前帧进行自动分割的递推式视频前景目标分割方法.目前视频分割主要有以下三大难点尚未被很好地解决,首先是前景与背景的颜色区分度过小,造成不易将其分割.其次,视频中频繁出现的局部遮挡与暴露区域,会破坏递推式视频分割的连续性,使得分割推理错误,造成结果不正确.最后,前/背景颜色模型所选用的采样方法,更是影响结果的重要因素之一.因此本文着重对此三类问题给出了相应的解决方案,利用前景物体具有局部运动一致性的特征,解决前/背景相似颜色问题.对于遮挡与暴露问题,通过自适应的局部窗口调整进行扩展采样,以修复其递推连续性.其次,本文使用了一种基于近似颜色块的采样方法,使得所建立的颜色模型更为明确干净,以减少不相关信息所带来的影响.最后,利用视频的时空一致性特征,合成出最终的分割结果.实验表明,本文所提出的递推式视频分割方法,能有效地解决以上三个问题,并且相比于其他方法,特别地对于复杂的视频测试场景,本文方法能获取更好的自动分割效果.     

The Munsell Color System: A scientific compromise from the world of art

Color systems make accurate color specification and matching possible in science, art, and industry by defining a coordinate system for all possible color perceptions. The Munsell Color System, developed by the artist Albert Henry Munsell in the early twentieth century, has influenced color science to this day. I trace the development of the Munsell Color System from its origins in the art world to its acceptance in the scientific community.Munsell's system was the first to accurately and quantitatively describe the psychological experience of color. By considering the problems that color posed for Munsell's art community and examining his diaries and published material, I conclude that Munsell arrived at his results by remaining agnostic as to the scientific definition of color, while retaining faith that color perceptions could be objectively quantified. I argue that Munsell was able to interest the scientific community in his work because color had become a controversial topic between physicists and psychologists. Parts of Munsell's system appealed to each field, making it a workable compromise. For contrast, I suggest that three contemporary scientists with whom Munsell had contact – Wilhelm Ostwald, Ogden Rood, and Edward Titchener – did not reach the same conclusions in their color systems because they started from scientific assumptions about the nature of color.     

Pierre-Joseph Macquer an Eighteenth-Century Artisanal-Scientific Expert

Summary

Pierre-Joseph Macquer (1718–1784) is well known as one of the major chemists in the eighteenth century as a theoretician and a teacher. He is also known for his works on dyeing. This paper presents a new face of Macquer. He proposed a theory on mordants in dyeing as early as 1775. Besides his activity at the Académie des sciences, he played an important role in Government as the commissioner of dyeing from 1766 where he established close links with artisan inventors. Académicien chimiste at the royal Manufactory of Sèvres from 1757, he was also the inventor of French porcelain. His notebooks show his organization, method, courage, passion and obstinacy in the search for the paste for hard porcelain. He also proposed an interpretation of its formation. Macquer was both a theoretician and a practical expert in dyeing as well as in porcelain making. He managed to bridge the gap between science and art.     

On a dynamic mixture GARCH model

This paper proposes a new mixture GARCH model with a dynamic mixture proportion. The mixture Gaussian distribution of the error can vary from time to time. The Bayesian Information Criterion and the EM algorithm are used to estimate the number of parameters as well as the model parameters and their standard errors. The new model is applied to the S&P500 Index and Hang Seng Index and compared with GARCH models with Gaussian error and Student's t error. The result shows that the IGARCH effect in these index returns could be the result of the mixture of one stationary volatility component with another non‐stationary volatility component. The VaR based on the new model performs better than traditional GARCH‐based VaRs, especially in unstable stock markets. Copyright © 2008 John Wiley & Sons, Ltd.