A robust regression approach for spectrophotometric signal analysis

The effectiveness of a regression method strongly depends on the characteristics of the considered regression problem. As a consequence, this makes it difficult to choose a priori the most appropriate algorithm for a given dataset. This issue is faced in this work through a novel regression approach based on the fusion of an ensemble of different regressors. In order to implement the proposed robust multiple system (RMS), four different fusion strategies are explored. In this context, we propose a novel fusion strategy named selection‐based strategy (SBS) that provides as output the estimate obtained by the regression algorithm (included in the ensemble) characterized by the highest expected accuracy in the region of the feature space associated with the considered model. The SBS is based not on a direct combination of the estimates yielded by all the regressors but on a selection mechanism that identifies the expected best available estimate. For such purpose, it exploits the accuracies of the regressors included in the ensemble in different portions of the input feature space. The experimental assessment of the RMS was carried out on three different datasets: a wine, an orange juice, and an apple datasets. The obtained experimental results suggest that, in general, the fusion of an ensemble of different regression algorithms leads to a regression process that is more robust and sometimes also more accurate than traditional regression methods. In particular, the proposed SBS method represents an effective solution to carry out the fusion process. Copyright © 2012 John Wiley & Sons, Ltd.     

基于RN和GN的两种RBF神经网络的工程造价预测模型

建立了调用NEWRB函数的正规化网络RN和基于K-means聚类的广义网络GN的两种RBF‘神经网络的工程造价预测模型,以55个厦门市工程造价案例进行实证分析.结果表明:当调用NEWRB函数构建RBF模型时,其性能主要取决于分布宽度,而基于K-means聚类的RBF神经网络主要取决于重叠系数和隐含层节点数;基于广义网络GN的RBF神经网络模型的训练效果较差,但学习速度更快、预测精度更高.     

基于RBF网络的有源噪声控制

提出一种神经自适应噪声有源控制（ANC）的方法。应用RBF（Radial Basis Function)网络对噪声进行有源控制。针对RBF的网络特点,使用递阶遗传算法确定网络参数（连接权、隐节点中心和宽度），同时解决了网络拓扑结构的优化训练。利用RBF网络的有源噪声控制系统用于三维空间传播的宽频带空调噪声的降噪获得了良好的效果。     

扰动情况下基于RBF网络的混沌系统同步

基于RBF网络,在扰动存在的情况下,对混沌系统的同步和异结构同步问题进行了研究.研究表明本方法能有效地克服干扰对同步所造成的破坏,实现良好的同步效果.最后,利用Lorenz系统进行了仿真.仿真结果证明了所给方法的有效性. 关键词： 混沌系统 同步 RBF神经网络     

AERODYNAMIC OPTIMIZATION OF HIGH-SPEED TRAIN BASED ON RBF MESH DEFORMATION

基于局部型函数三维参数化方法、改进的蚁群算法和改进的克里金（Kriging）代理模型,开展了列车头型的三维气动减阻优化设计研究。为了避免复杂几何外形大变形情况下千万量级网格的重复生成,提高高速列车头型优化设计的效率,引入了缩减控制点的径向基函数网格变形技术。优化结果表明：径向基函数网格变形技术在不降低网格质量的情况下可以有效缩短网格变形的时间消耗,能够用于复杂几何外形的气动优化设计;在给定的设计空间内,控制鼻锥外形的6个关键设计参数对列车气动阻力的影响呈单调递增关系;优化后,在满足约束条件的情况下,简化外形列车的整车气动阻力减小5.41%,头尾车减阻效果明显,中间车气动阻力基本不变。     

一种基于RBF神经网络的打印机光谱预测模型

本文提出一种基于RBF(Radial Basis Function,径向基函数)神经网络的打印机光谱预测模型,通过扩展神经网络模型输入变量的项数提高模型的预测精度,扩展项多采用通道驱动值的交叉值、平方值.实验结果表明[1cmy]项的引入能够有效提高模型的预测精度,同时提高网络的泛化能力.而引入[cm² cy² mc² my² yc² ym²]项会导致模型预测精度以及泛化能力降低.[1 cmy]、[c²m²y²]和[cm cy my]项的组合在预测精度和模型泛化能力上均是最优化的,对总样本预测的色度精度为0.475ΔE₀₀,光谱精度RMSE为0.43%.因此选择[1 cmy c²m²y² cm cy my c m y]作为输入变量的RBF神经网络训练模型是满足高精度光谱预测的最优模型.     

基于径向基神经网络的白矮主序双星自动分类

白矮主序双星的光谱特征是决定其类型的关键因素,如何有效提取恒星光谱的特征是亟待解决的问题。提出一种新的方法,通过构建模型捕获恒星光谱数据的特征,对SDSS-DR10海量光谱进行自动分类。径向基神经网络作为一种有效的计算模型,在数值逼近和目标分类上均有较好的表现效果,但由于目前神经网络超参数的确定大多数依赖于实验经验,很大程度上制约了算法能力的发挥。在分析白矮主序双星光谱数据的高维分布特征的基础上,提出一种基于径向基神经网络的白矮主序双星自动分类模型,并以白矮主序双星的光谱特征为导向,针对恒星光谱提出了中心准则和宽度准则以确定神经网络的超参数,大幅度提高了模型的准确度。实验对分类模型进行数值训练并使用训练的模型对SDSS-DR10光谱数据进行白矮主序双星的自动分类,共发现4 631个白矮主序双星,通过Simbad,NED和Google交叉验证后发现其中有25个是未予以收录的新候选体。实验结果验证了该模型在大规模白矮主序双星自动分类任务中的有效性,新发现的白矮主序双星也为特殊天体的进一步研究补充了有效数据。     

压电驱动快速反射镜的自适应反演滑模控制

由压电驱动器驱动的快速反射镜(FSM)广泛应用于各种精密稳定跟踪系统,FSM的控制精度决定了系统的跟踪精度。但压电驱动器具有严重的迟滞非线性干扰,针对这一缺点,应用自适应径向基RBF神经网络对迟滞干扰进行非线性逼近,并在此基础上结合滑模控制和反演法,设计了自适应反演滑模(ABSM)控制器。仿真实验表明,相对于滑模控制器,ABSM控制器的最大跟踪误差和均方根误差为分别降低了57.26%和52.53%,提高了FSM的控制精度。     

Extending the working calibration ranges of four hexachlorocyclohexane isomers in gas chromatography-electron capture detector by radial basis function neural network

A radial basis function neural network (RBFNN) method was developed for the first time to model the nonlinear calibration curves of four hexachlorocyclohexane (HCH) isomers, aiming to extend their working calibration ranges in gas chromatography-electron capture detector (GC-ECD). Other 14 methods, including seven parametric curve fitting methods, two nonparametric curve fitting methods, and five other artificial neural network (ANN) methods, were also developed and compared. Only the RBFNN method, with logarithm-transform and normalization operation on the calibration data, was able to model the nonlinear calibration curves of the four HCH isomers adequately. The RBFNN method accurately predicted the concentrations of HCH isomers within and out of the linear ranges in certified test samples. Furthermore, no significant difference (p > 0.05) was found between the results of HCH isomers concentrations in water samples calculated with RBFNN method and ordinary least squares (OLS) method (R² > 0.9990). Conclusively, the working calibration ranges of the four HCH isomers were extended from 0.08-60 ng/ml to 0.08-1000 ng/ml without sacrificing accuracy and precision by means of RBFNN. The outstanding nonlinear modeling capability of RBFNN, along with its universal applicability to various problems as a “soft” modeling method, should make the method an appealing alternative to traditional modeling methods in the calibration analyses of various systems besides the GC-ECD.     

$L^p(K)$中RBF神经网络的系统识别问题

L^p approximation problems in system identification with RBF neural networks are investigated. It is proved that by superpositions of some functions of one variable in L^ploc（R）, one can approximate continuous functionals defined on a compact subset of L^P（K） and continuous operators from a compact subset of L^p1 （K1） to a compact subset of L^p2 （K2）. These results show that if its activation function is in L^ploc（R） and is not an even polynomial, then this RBF neural networks can approximate the above systems with any accuracy.