Prediction of thermal conductivity of polymer-based composites by using support vector regression

Support vector regression (SVR) combined with particle swarm optimization (PSO) for its parameter optimization, was proposed to establish a model to predict the thermal conductivity of polymer-based composites under different mass fractions of fillers (mass fraction of polyethylene (PE) and mass fraction of polystyrene (PS)). The prediction performance of SVR was compared with those of other two theoretical models of spherical packing and flake packing. The result demonstrated that the estimated errors by l...     

Estimating leisure noise in Spanish cities

Noise pollution is a great environmental problem in big cities. In particular, the leisure noise is one of the most significant problems during the night period and the responsible of a growing number of complaints and sleep disturbances from population.For that reason, with the aim of approaching an analysis on spatio-temporal evolution of the leisure noise, some areas with leisure activity as main noise source was sampled. With these results, a protocol to address an appropriate action against leisure noise problems is put forth. This procedure explains how to diagnose a leisure area or predict the noise levels inside these areas. Therefore, the proposed procedure could help taking better solutions in order to reduce the impact of leisure noise on the exposed population.     

Physical implication of two problems in transition prediction of boundary layers based on linear stability theory

Up to now,the most widely used method for transition prediction is the one based on linear stability theory.When it is applied to three-dimensional boundary layers,one has to choose the direction,or path,along which the growth rate of the disturbance is to be integrated.The direction given by using saddle point method in the theory of complex variable function is seen as mathematically most reasonable.However,unlike the saddle point method applied to water waves,here its physical meaning is not so obvious,as the frequency and wave number may be complex.And on some occasions,in advancing the integration of the growth rate of the disturbance,up to a certain location,one may not be able to continue the integration,because the condition for specifying the direction set by the saddle point method can no longer be satisfied on the basis of continuously varying wave number.In this paper,these two problems are discussed,and suggestions for how to do transition prediction under the latter condition are provided.     

TargetATPsite: A template‐free method for ATP‐binding sites prediction with residue evolution image sparse representation and classifier ensemble

Understanding the interactions between proteins and ligands is critical for protein function annotations and drug discovery. We report a new sequence‐based template‐free predictor (TargetATPsite) to identify the Adenosine‐5′‐triphosphate (ATP) binding sites with machine‐learning approaches. Two steps are implemented in TargetATPsite: binding residues and pockets predictions, respectively. To predict the binding residues, a novel image sparse representation technique is proposed to encode residue evolution information treated as the input features. An ensemble classifier constructed based on support vector machines (SVM) from multiple random under‐samplings is used as the prediction model, which is effective for dealing with imbalance phenomenon between the positive and negative training samples. Compared with the existing ATP‐specific sequence‐based predictors, TargetATPsite is featured by the second step of possessing the capability of further identifying the binding pockets from the predicted binding residues through a spatial clustering algorithm. Experimental results on three benchmark datasets demonstrate the efficacy of TargetATPsite. © 2013 Wiley Periodicals, Inc.     

Nonlinear analysis of near-wall turbulence time series

We present the results of the nonlinear analysis of a near-wall turbulence signal, measured using a Laser Doppler Anemometer. Despite being preliminary in part, the results highlight some interesting aspects of the dynamics of coherent structures. Careful reconstruction of the dynamics from the scalar time series highlight particular phases of the bursting cycle, corresponding to the VITA detections, immersed in the overall high-dimension dynamics: the possibility to distinguish these trajectories in the embedding space provides indications for conditional sampling techniques on the reconstructed attractor. The application of nonlinear prediction to a system of high-dimension produces forecasts of rapidly decreasing quality over time, with no consequences for practical applications. Much more interesting instead is the analysis of the forecast errors for very short forecast intervals, where we observed a greater difficulty to predict the beginning of coherent structure instability probably due to the quasi-random forcing of the outer flow.     

非等距时间序列的GM（1，N）模型及其在地下工程中的应用

文献〔１〕给出了等距时间序列的灰色预测ＧＭ（１，Ｎ）模型，本文把它推广到非等距时间序列，得到了相应模型，并应用于地下工程的几个实例     

基于数据驱动的卫星锂离子电池寿命预测方法

锂离子电池由于具有工作电压高、体积小、质量轻、比能量高、寿命长和自放电率小等优点,成为 替代传统镍氢、镍镉电池的第三代航天器用储能电源。寿命预测是锂离子电池健康管理的重要方面,是掌 握电源衰退的重要手段。锂离子电池剩余使用寿命预测问题已成为电子系统健康管理领域的研究热点和具 有挑战性的问题之一。本文基于NASA埃姆斯中心的锂离子电池地面试验采集数据,首先分析了3种类型 的锂离子电池预测方法,之后重点研究了几种有效的数据驱动的锂离子电池寿命预测方法,并对各种预测 方法的效果进行了评价。实验结果表明,本文提出的方法能够有效的用于基于数据驱动的锂离子电池寿命 预测中,具有较强的工程应用价值。     

面向数据挖掘的静态电源综合故障诊断研究

静态电源是机场桥载设备最重要的组成部件之一,应用非常广泛。然而,其产生的频发故障会造成设备利用率低、修复率时长和经济损失等问题,在基于桥载设备的安监系统上,提出了静态电源综合故障诊断方法,通过数据挖掘软件,建立了静态电源故障诊断预测模型。通过在线数据库测试结果表明得到了综合故障诊断方法在预测静态电源故障上显现的特点,得到了静态电源的未来状态,实现了对静态电源的实时故障进行预测,进而为解决故障提供方向和目标,最终达到降低经济损失最大化的目的。     

语音信号序列的Volterra预测模型

对给定的英语音素、单词和语句进行了采集并完成预处理. 分别应用互信息法和Cao 氏法确定了实际采集的语音信号序列的延迟时间和嵌入维数, 以完成语音序列的相空间重构. 通过计算实际采集的语音信号序列的最大Lyapunov指数, 完成了语音信号的混沌特性识别, 判定其具有混沌特性. 引入Volterra级数, 提出了一种具有显式结构的语音信号非线性预测模型. 为克服最小均方误差算法在Volterra模型系数更新时固有的缺点, 在最小二乘法基础上, 应用基于后验误差假设的可变收敛因子技术, 构建了一种基于Davidon-Fletcher-Powell算法的二阶Volterra 模型(DFPSOVF), 并将其应用于具有混沌特性的语音信号序列预测. 仿真结果表明: DFPSOVF非线性预测模型对于单帧和多帧语音信号均具有更好的预测精度, 优于线性预测模型, 并且能够很好地反映语音序列变化的趋势和规律, 完全可以满足语音预测的要求; 可以根据语音信号序列的嵌入维数选取预测模型的记忆长度. 所提出模型可以为语音信号重构和压缩编码开辟一条新途径, 以改善语音信号处理方法的复杂度和处理效果.     

基于核各向异性扩散的红外小目标检测

为了减少红外图像中背景边缘对检测的影响,提出了一种具有鲁棒性的弱小目标检测算法,该算法利用核各向异性扩散模型进行背景预测,再与原图像差分实现弱小目标检测。为了提高算法的自适应能力,提出了一种鲁棒性扩散系数,能够根据图像背景的起伏程度自适应调整扩散系数曲线的陡峭程度。实验结果表明,与现有的检测算法相比,该算法能够在不同类型的复杂背景下有效抑制背景及其边缘,保留目标大小,降低虚警率,具有更强的鲁棒性。