用三层网络逼近任何连续函数

在神经网络基础理论中，有一个这样的问题：怎样用三层网络实现对任意连续函数的映射，本文利用Ｋｏｈｏｎｅｎ提出的Ｓｅｌｆ－ＯｒｇａｎｉｚｉｎｇＦｅａｔｕｒｅＭａｐ竞争型网络以及区域的２Ｍ－分割概念，解决了用具有有限个隐层单元的三层网络以任意精度逼近一个连续函数的问题，具体的网络结构和学习算法已给出．     

Inferring Invasion History of Red Swamp Crayfish (Procambarus clarkii) in China from Mitochondrial Control Region and Nuclear Intron Sequences

Identifying the dispersal pathways of an invasive species is useful for adopting the appropriate strategies to prevent and control its spread. However, these processes are exceedingly complex. So, it is necessary to apply new technology and collect representative samples for analysis. This study used Approximate Bayesian Computation (ABC) in combination with traditional genetic tools to examine extensive sample data and historical records to infer the invasion history of the red swamp crayfish, Procambarus clarkii, in China. The sequences of the mitochondrial control region and the proPOx intron in the nuclear genome of samples from 37 sites (35 in China and one each in Japan and the USA) were analyzed. The results of combined scenarios testing and historical records revealed a much more complex invasion history in China than previously believed. P. clarkii was most likely originally introduced into China from Japan from an unsampled source, and the species then expanded its range primarily into the middle and lower reaches and, to a lesser extent, into the upper reaches of the Changjiang River in China. No transfer was observed from the upper reaches to the middle and lower reaches of the Changjiang River. Human-mediated jump dispersal was an important dispersal pathway for P. clarkii. The results provide a better understanding of the evolutionary scenarios involved in the rapid invasion of P. clarkii in China.     

用于综合评判的一种权重计算方法

在综合评判和决策过程中，属性权重的确定是其中一个重要环节。由于常用方法存在诸多缺点,为此根据信息具有粒度的思想，基于粗糙集两个近似精度来科学定义属性的重要性，并将之作为综合评判中的客观权重。此外考虑到客观经验因素，将客观扭值和主观权重相结合，最后得到有利于正确评价和决策的综合权重。流型识别仿真试验结果表明：文章提出的权重计算方法较为完备，其应用效果优于其它权重计算方法。     

“λ”形近似直线机构稳健性能分析

以“λ”形机构为研究对象,在建立机构综合模型基础上,提出取实际机构输出的连杆曲线直线近似度与理想直线近似度之间的偏差作为质量特性指标的稳健性能评价模型,结合示例计算并绘制满足输出精度要求的概率曲线图和机构稳健性能图,得到了满足约束条件的局部最优和全局最优稳健机构解,设计者可以很容易选取稳健性能优越的“λ”形机构.     

基于SPEA-Ⅱ算法的网络多层次安全访问控制方法

当前网络技术迅猛发展,网络安全问题日益突出,尤其是访问控制安全性。为了解决目前网络系统中存在的威胁检测精度和安全访问控制问题,引入SPEA-II算法,提出一种新的网络多层次安全访问控制方法。深入分析网络多层次访问控制机制原理,明确安全访问控制目标。结合网络边界区域、传输信道区域、移动终端设备区域,建立边界访问控制目标函数、安全威胁检测目标函数、用户身份认证目标函数,提高网络系统安全威胁检测精度。利用SPEA-Ⅱ算法对联合目标函数进行求解,获取网络多层次安全访问控制机制最佳方案。通过迭代逐渐接近近似最优解集,建立最优安全策略组合,计算攻击流量因子完成网络多层次安全访问控制。实验结果表明,所提方法的网络多层次威胁检测精度为94%,安全评估为0.96。由此证明,所提方法的网络多层次安全访问控制效果较好,具有较强的安全性和适应性,能够为中移互联网领域的网络安全策略提供技术支持。     

Evaluating process centering with large samples – an approximate lower bound on the accuracy index

Process accuracy index C_a has been proposed in the manufacturing industry to provide numerical measures on process centering (the ability to cluster around the specification mid-point). Investigations on process accuracy in existing engineering statistics and quality assurance literature are all focused on small samples with a normality assumption. In this paper, a natural estimator of process accuracy index is considered. The limiting distribution and related large sample properties of the considered estimator are studied under general populations having fourth central moment exists. A decision-making procedure based on an approximate 100(1-α)% lower bound of the process accuracy index is also constructed. A practical example is demonstrated to illustrate how the proposed procedure may be applied for in-plant applications to judge whether the process runs under the desirable accuracy requirement.     

Software support for the construction of syntactic pattern recognisers by a research engineer

Syntactic pattern recognition is introduced and it is suggested that a research engineer will probably be less familiar with the language theory underlying syntactic pattern recognition than with the statistical ideas connected with decision theoretic methods. For this reason application of syntactic pattern recognition will only develop if software support is provided. The steps of grammatical inference, recogniser construction and recogniser optimisation are outlined. They are included in a scheme by which, working interactively and iteratively, a syntactic pattern recogniser can be produced. Some comments on an implementation are given.     

Optimum Pareto design of non-linear predictive control with multi-design variables for PEM fuel cell

The control process of Proton Exchange Membrane Fuel Cells (PEMFCs) is a difficult task due to the non-linearities and uncertainties associated with the electrochemical processes governing it. Designing of a non-linear controller based on model predictive control for PEMFCs is presented in some previous works to regulate the cell voltage or power output based on just one of the input variables like hydrogen pressure or operating temperature respectively, but they use a constant signal for other important control variables due to computational limitations in on-line optimization of multivariable highly non-linear system. In this paper, by use of Approximate Predictive Control (APC) method based on neural network model, the whole control process is designed with three input variables. Operating temperature and hydrogen pressure are assumed as control design variables and current density as measured disturbance to manage the cell voltage. Moreover, multi-objective optimization based on multi-objective uniform-diversity genetic algorithm (MUGA) is used for optimal selection of the parameters of controller. The comparison of the obtained results with those in literature demonstrates the superiority of the results of this work.