Directed searching optimization algorithm for constrained optimization problems

A directed searching optimization algorithm (DSO) is proposed to solve constrained optimization problems in this paper. The proposed algorithm includes two important operations — position updating and genetic mutation. Position updating enables the non-best solution vectors to mimic the best one, which is beneficial to the convergence of the DSO; genetic mutation can increase the diversity of individuals, which is beneficial to preventing the premature convergence of the DSO. In addition, we adopt the penalty function method to balance objective and constraint violations. We can obtain satisfactory solutions for constrained optimization problems by combining the DSO and the penalty function method. Experimental results indicate that the proposed algorithm can be an efficient alternative on solving constrained optimization problems.     

Effects of the number of hidden nodes used in a structured-based neural network on the reliability of image classification

A structured-based neural network (NN) with backpropagation through structure (BPTS) algorithm is conducted for image classification in organizing a large image database, which is a challenging problem under investigation. Many factors can affect the results of image classification. One of the most important factors is the architecture of a NN, which consists of input layer, hidden layer and output layer. In this study, only the numbers of nodes in hidden layer (hidden nodes) of a NN are considered. Other factors are kept unchanged. Two groups of experiments including 2,940 images in each group are used for the analysis. The assessment of the effects for the first group is carried out with features described by image intensities, and, the second group uses features described by wavelet coefficients. Experimental results demonstrate that the effects of the numbers of hidden nodes on the reliability of classification are significant and non-linear. When the number of hidden nodes is 17, the classification rate on training set is up to 95%, and arrives at 90% on the testing set. The results indicate that 17 is an appropriate choice for the number of hidden nodes for the image classification when a structured-based NN with BPTS algorithm is applied.     

面向Lustre文件系统的固态盘存储加速技术研究

磁盘的随机I/O延时制约了存储系统的性能提高,具备高性能随机I/O特性的固态盘(solid state disk,SSD)逐渐成为关注的热点.分析了磁盘、Flash型SSD以及DRAM型SSD三类设备不同的性能特点,讨论了SSD存储加速技术的研究现状,提出了一种面向Lustre文件系统的固态盘存储加速系统架构,介绍和分析了各模块的构成与原理,提出了对象迁移策略.     

Optimization of the profile of the nano-mold and the imprinting conditions by numerical simulation method

The profile of the nano-mold can also influence the replication precision of the final 2D polymer nanochannels. In the present work, the aspect ratio of the nano-mold, the duty ratio of the nano-mold, and the thickness ratio were investigated by numerical simulation method. The replication precision of the 2D polymer nanochannels is evaluated by imprinting ratio. The simulation results show that the lower aspect ratio and duty ratio is associated with higher imprinting ratio. However, the imprinting ratio increases with the thickness ratio at beginning and then keeps nearly unchanged as the thickness ratio continuously increases. The proper thickness ratio is two. The imprinting temperature and pressure were also optimized according the imprinting ratio. The simulation results show that the optimal imprinting temperature and pressure were 120 °C and 1.5 MPa, respectively.     

L1-norm based null space discriminant analysis

Multimedia Tools and Applications - Null space based linear discriminant analysis (NSLDA) is a well-known feature extraction method, which can make use of the most discriminant information in the...     

Analyses and improvement of a broadcasting multiple blind signature scheme based on quantum GHZ entanglement

A broadcasting multiple blind signature scheme based on quantum GHZ entanglement has been presented recently. It is said that the scheme’s unconditional security is guaranteed by adopting quantum key preparation, quantum encryption algorithm and quantum entanglement. In this paper, we prove that each signatory can get the signed message just by an intercept–resend attack. Then, we show there still exist some participant attacks and external attacks. Specifically, we verify the message sender Alice can impersonate each signatory to sign the message at will, and so is the signature collector Charlie. Also, we demonstrate that the receiver Bob can forge the signature successfully, and with respect to the external attacks, the eavesdropper Eve can modify the signature at random. Besides, we discover Eve can change the signed message at random, and Eve can impersonate Alice as the message sender without being discovered. In particular, we propose an improved scheme based on the original one and show that it is secure against not only the attacks mentioned above but also some collusion attacks.     

Mean square detectability of LTI systems over finite-state digital block-fading channels

This paper studies the mean square quadratic (MSQ) detectability for multi-output networked systems over finite-state digital block-fading channels. The packet-loss rate of each digital fading channel depends on the channel power gain, as well as packet length and power level used for transmission. A finite-state random process is introduced to model time-varying fading channels, which characterizes various configurations of physical communication environment and/or different channel fading amplitudes. Necessary and sufficient conditions for MSQ detectability over finite-state Markov digital block-fading channels are given in the form of algebraic Riccati equations or linear matrix inequalities (LMIs). The estimation gain is given as a function of estimated/observed channel state. In addition, explicit conditions on network for MSQ detectability over finite-state independent identically distributed (i.i.d.) digital block-fading channels are presented in terms of the unstable poles of the multi-output plant. Finally, an application to Gilbert-Elliott channels (GECs) is provided to demonstrate the derived results.     

Neural-network-based fuzzy logic control of a 3D rigid pendulum

International Journal of Control, Automation and Systems - This paper concerns the attitude control problem of a 3D rigid pendulum based on dynamic T-S fuzzy neural model. A generalized 3D rigid...     

一种基于颜色特征的感兴趣目标提取方法

针对现有的感兴趣区域(ROI)提取方法边缘不清晰、区域不完整等问题,提出一种ROI提取方法.首先采用颜色局部特征的信息量大小衡量兴趣度的大小,然后融合颜色特征信息量图获得图像的显著图(SM),再进行阈值分割,得到显著值大的区域,即ROI.实验结果表明,该方法能有效地提取出感兴趣的对象,SM中目标区域的显著值均匀、边缘清晰;与人工标记的ROI比较,该方法召回率为79.71％,精度为78.53％,优于已有的ROI提取方法.     

网页设计与制作教学网站的设计与实现

为了提高教学效率,使优秀的教学资源和信息能够在更大范围、短时间内迅速传播,采用Dreamweaver工具和ASP技术开发了网络教学平台.教学网站前台采用动态网页设计,后台使用SQL Server 2005数据库以满足大用户量的需求.网站实现了在线答疑、讨论、课件上传下载、教学管理及用户管理等功能,并通过了测试和试运行.该网站的实现为教师和学生提供了一个高效、方便的网络教学环境.