结合多尺度空间滤波和层级网络的基于结构保持的高光谱特征选择

为了充分利用高光谱图像蕴含的丰富的光谱信息和空间信息,提出了结合多尺度空间滤波和层级网络的基于结构保持的高光谱特征选择算法.算法利用基于l2,1范数的数学模型,选出同时保存全局相似性结构和局部流形结构的特征子集;在多个尺度的窗口中使用双边滤波,自适应计算滤波核,自动在光谱数据中融入空间信息,增强了类内相似性和类间相异性,避免了参量选择;引入层级结构实现空间信息和光谱信息的深入融合,提高了分类准确度;讨论了层级数目和窗口尺度个数对分类准确度的影响.在Indian Pines和PaviaU两个数据集的实验表明,该算法在大部分地物种类上的分类准确度都有较大幅度的提升,总体分类准确度分别达到90.98%和94.20%,相比其他方法明显提高了地物分类准确度.     

张量重正化群方法及其应用

张量重正化群方法是近年来发展起来的一种新的数值计算方法,它将经典配分函数和量子波函数的张量网络表示与重正化群方法相结合,在强关联系统的数值研究中,发挥着越来越重要的作用。文章以经典统计模型和量子格点模型为例,简要介绍了张量重正化群的一些基础知识和研究给定物理模型的一般性思路,并对张量重正化群未来可能的发展方向和亟待解决的问题进行了讨论。     

Temperature dependence of heat conduction coefficient in nanotube/nanowire networks

Studies on heat conduction are so far mainly focused on regular systems such as the one-dimensional(1D) and twodimensional(2D) lattices where atoms are regularly connected and temperatures of atoms are homogeneously distributed.However, realistic systems such as the nanotube/nanowire networks are not regular but heterogeneously structured, and their heat conduction remains largely unknown. We present a model of quasi-physical networks to study heat conduction in such physical networks and focus on how the network structure influences the heat conduction coefficient κ. In this model,we for the first time consider each link as a 1D chain of atoms instead of a spring in the previous studies. We find that κ is different from link to link in the network, in contrast to the same constant in a regular 1D or 2D lattice. Moreover, for each specific link, we present a formula to show how κ depends on both its link length and the temperatures on its two ends.These findings show that the heat conduction in physical networks is not a straightforward extension of 1D and 2D lattices but seriously influenced by the network structure.     

A generalized model of TiO_x-based memristive devices and its application for image processing

Memristive technology has been widely explored, due to its distinctive properties, such as nonvolatility, high density,versatility, and CMOS compatibility. For memristive devices, a general compact model is highly favorable for the realization of its circuits and applications. In this paper, we propose a novel memristive model of TiO_x-based devices, which considers the negative differential resistance(NDR) behavior. This model is physics-oriented and passes Linn's criteria. It not only exhibits sufficient accuracy(IV characteristics within 1.5% RMS), lower latency(below half the VTEAM model),and preferable generality compared to previous models, but also yields more precise predictions of long-term potentiation/depression(LTP/LTD). Finally, novel methods based on memristive models are proposed for gray sketching and edge detection applications. These methods avoid complex nonlinear functions required by their original counterparts. When the proposed model is utilized in these methods, they achieve increased contrast ratio and accuracy(for gray sketching and edge detection, respectively) compared to the Simmons model. Our results suggest a memristor-based network is a promising candidate to tackle the existing inefficiencies in traditional image processing methods.     

Recursion-transform method and potential formulae of the m × n cobweb and fan networks

In this paper, we made a new breakthrough, which proposes a new recursion–transform(RT) method with potential parameters to evaluate the nodal potential in arbitrary resistor networks. For the first time, we found the exact potential formulae of arbitrary m × n cobweb and fan networks by the RT method, and the potential formulae of infinite and semi-infinite networks are derived. As applications, a series of interesting corollaries of potential formulae are given by using the general formula, the equivalent resistance formula is deduced by using the potential formula, and we find a new trigonometric identity by comparing two equivalence results with different forms.     

基于自编码网络流形学习的毛竹笋不溶性膳食纤维含量红外光谱建模

提出了一种结合自编码网络(AN)流形学习和偏最小二乘(PLS)法的红外光谱建模方法AN-PLS。AN-PLS方法首先用AN算法对红外光谱数据进行非线性降维,再结合PLS建立回归模型。利用该方法建立了毛竹笋中不溶性膳食纤维含量的近红外光谱和中红外光谱回归模型。结果表明,用AN-PLS方法建立的回归模型,比用其他常用光谱数据预处理方法结合PLS及用单独PLS算法建立的模型具有更小的预测均方根误差RMSEP和更高的决定系数R2,因此,AN-PLS具有较优的建模与预测能力,利用近红外光谱和中红外光谱技术结合AN-PLS建模,可实现毛竹笋中不溶性膳食纤维含量的准确测量。     

智能电网的10G-EPON中基于贝叶斯分类的业务感知机制

随着智能电网的发展及其多种信息业务的涌现,10G-EPON作为业务接入技术日益成为重要支撑;然而业务的多元化对10G-EPON的多业务支撑能力提出了重要挑战.为了适应电力系统中多种不同类型业务的需求,本文对智能电网的信息业务特性进行分析,提出了一种基于贝叶斯分类的10G-EPON业务感知机制;并且根据10G-EPON中OLT与ONU的主从式网络架构特点,提出了业务感知的主从式实现方式.该机制使用贝叶斯网络分析数据包的特征,进而确认待传送业务的类型.在贝叶斯业务分类的基础上,通过OLT和ONU之间的交互决定业务的资源分配和传输策略.为了验证新机制的有效性,分别从时延和丢包率两方面进行系统仿真.仿真结果表明,所提出的基于贝叶斯分类的业务感知机制在时延和丢包率具有显著的优势,能够实现业务与10G-EPON的高效匹配,提高10G-EPON在智能电网应用中多业务的区分支持能力.     

Rhythm dynamics of complex neuronal networks with mixed bursting neurons

The spatiotemporal order and rhythm dynamics of a complex neuronal network with mixed bursting neurons are studied in this paper. A quantitative characteristic, the width factor, is introduced to describe the rhythm dynamics of an individual neuron, and the average width factor is used to characterize the rhythm dynamics of a neuronal network. An r parameter is introduced to denote the ratio of the short bursting neurons in the network. Then we investigate the effect of the ratio on the rhythm dynamics of the neuronal network. The critical value of r is derived, and the neurons in the network always remain short bursting when the r ratio is larger than the critical value.     

量子信令交换机模型设计及性能分析

本文提出了量子信令交换机的模型, 该交换机由经典信息控制模块、交换控制模块和量子交换模块三部分组成. 经典控制模块负责将纠缠初态信息传送给纠缠测量及交换单元并更新路由信息. 交换控制模块实现通路选择, 为纠缠对的分发提供通路. 量子交换模块制备纠缠对, 进行Bell态的测量, 完成纠缠交换. 量子信令交换机可以实现多用户间的信令传输及局域网通信. 通过对交换机的性能分析与仿真, 结果表明该交换机结构简单、安全保密、便于扩展、时延小, 对于构建量子通信网有很好的支撑作用. 关键词： 量子通信 量子信令网 量子信令交换机 纠缠交换     

EMD与神经网络在气液两相流流型识别中的应用

本文提出了EMD与Elman神经网络相结合的气液两相流流型识别的新方法.将压差波动信号经验模态分解(EMD)后的固有模态函数(IMF)进行分析、提取IMF能量作为Elman神经网络的输入特征向量,对水平管内的气液两相流流型进行识别.实验结果表明:该方法优于BP网络且稳定、识别率高,具有可行性.