具有时滞的Cohen—Grossberg神经网络的Hopf分支全局存在性研究

摘要：利用泛函微分方程的度理论,研究一类具有时滞的Cohen-Grossberg神经网络的全局分支的存在性,研究结果为该类神经网络的应用设计提供理论基础．     

Strategic intelligence during coin detention operations – relational data and understanding latent terror networks

One aspect of the global “War on Terror” that has received limited coverage in the academic literature is the problem of detained persons as it relates to intelligence. This is a surprising oversight, given the number of detainees that the USA has been responsible for (over 25,000 were in custody in Iraq alone at one time during its peak). The detention environment offers a prime strategic intelligence opportunity for the US intelligence community to study the tactics and organizations of individuals who have been removed from the overall conflict. In this article an easily implemented collection program is recommended to be deployed in US/coalition detention centers. The primary recommendation is to gather relational data on detainee communication, both authorized and illicit, and to use these data to perform network analyses of terrorist groups and their individual members.     

Coordination mechanisms for a three‐stage reverse supply chain to increase profitable returns

In this article, we consider a generic electronic product that can be remanufactured or recycled at the end of its life cycle to generate new profit. We first describe the product return process and then present a customer segmentation model to capture consumers' different behaviors with respect to product return so that the retailer can work more effectively to increase the return volume. In regard to the collaboration between the retailer and the manufacturer, we explore a revenue‐sharing coordination mechanism for achieving a win‐win outcome. The optimality and sensitivity of the critical parameters in four strategies are obtained and examined both theoretically and numerically, which generate insights on how to manage an efficient consumer‐retailer‐manufacturer reverse supply chain, as well as on the feasibility of simplifying such a three‐stage chain structure. © 2012 Wiley Periodicals, Inc. Naval Research Logistics, 2013     

一种基于模糊竞争网络的主动声纳目标分类器

提出了一种基于模糊竞争网络的主动声纳目标分类器。该分类器结构简单,学习速度快,仅用少量样本进行训练便可获得良好的推广性能．便于实时学习和应用。通过对模式属于各类别的隶属度的分析,可以知道判决的可靠性,具有较好的应用价值。     

Analysis of algorithms for two‐stage flowshops with multi‐processor task flexibility

In this article we introduce a 2‐machine flowshop with processing flexibility. Two processing modes are available for each task: namely, processing by the designated processor, and processing simultaneously by both processors. The objective studied is makespan minimization. This production environment is encountered in repetitive manufacturing shops equipped with processors that have the flexibility to execute orders either individually or in coordination. In the latter case, the product designer exploits processing synergies between two processors so as to execute a particular task much faster than a dedicated processor. This type of flowshop environment is also encountered in labor‐intensive assembly lines where products moving downstream can be processed either in the designated assembly stations or by pulling together the work teams of adjacent stations. This scheduling problem requires determining the mode of operation of each task, and the subsequent scheduling that preserves the flowshop constraints. We show that the problem is ordinary NP‐complete and obtain an optimal solution using a dynamic programming algorithm with considerable computational requirements for medium and large problems. Then, we present a number of dynamic programming relaxations and analyze their worst‐case error performance. Finally, we present a polynomial time heuristic with worst‐case error performance comparable to that of the dynamic programming relaxations. © 2003 Wiley Periodicals, Inc. Naval Research Logistics, 2004.     

Waypoint analysis for command and control

Command and Control (C2) in a military setting can be epitomized in battles‐of‐old when commanders would seek high ground to gain superior spatial‐temporal information; from this vantage point, decisions were made and relayed to units in the field. Although the fundamentals remain, technology has changed the practice of C2; for example, enemy units may be observed remotely, with instruments of varying positional accuracy. A basic problem in C2 is the ability to track an enemy object in the battlespace and to forecast its future position; the (extended) Kalman filter provides a straightforward solution. The problem changes fundamentally if one assumes that the moving object is headed for an (unknown) location, or waypoint. This article is concerned with the new problem of estimation of such a waypoint, for which we use Bayesian statistical prediction. The computational burden is greater than an ad hoc regression‐based estimate, which we also develop, but the Bayesian approach has a big advantage in that it yields both a predictor and a measure of its variability. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2004     

未知非线性系统的神经网络建模与控制仿真研究

针对一类未知非线性系统的建模与控制问题,采用基于神经网络的多步预测控制算法进行了仿真研究,仿真对象选取控制工程界常用的单摆试验装置.该算法的实现步骤为:获取系统开环试验数据;辨识神经网络正向动态模型;设计非线性优化控制律.仿真研究结果表明,所提出的建模与控制方法是有效的.     

具有变时滞的离散型随机BAM神经网络的稳定性

研究一类具有变时滞的离散型随机BAM神经网络。通过构造Lyapunov泛函以及线性矩阵不等式（LMI）,得出了离散系统全局渐近稳定的充分条件。     

基于弱多径覆盖的虚电路恢复策略研究

虚电路技术可以为Ad Hoc网络中的节点提供面向连接的服务。然而,由于拓扑动态性等特征,Ad Hoc网络中的虚电路容易断开,从而导致业务时延、时延抖动增大甚至通信中断等问题。为此文章提出了一种基于弱多径覆盖的虚电路恢复策略,通过将虚电路机制与弱多径覆盖多径策略相结合,可以有效地减少虚电路恢复时间和需要重新预约资源的节点的数量。理论分析和仿真结果表明了该方法的有效性。     

基于小波包变换和BP网络的齿轮箱故障诊断

利用小波包分析提取出齿轮箱在各种工况下振动信号的有效成分,根据时频域特点提取比较典型的特征参量,利用改进的BP网络进行训练,根据训练结果判别齿轮箱的故障状态。