基于神经网络的离散变结构控制系统

系统研究了基于神经网络的离散变结构控制系统设计方法,提出了几种具体设计方案.神经网络的引入可以使滑模(变结构)控制具备学习与自适应能力,使控制信号得以柔化,从而能够减轻或避免困扰常规滑模控制器的抖振现象,改善控制效果.     

C3I系统仿真及仿真环境研究

首先讨论了C3I系统开发中存在的问题并提出解决这些问题的重要性,然后从C3I系统描述模型和仿真环境两个方面,介绍了国内外对于C3I系统仿真的研究现状和发展趋势.最后介绍了我们正在研制的基于对象Petri网的仿真演示环境的特点以及设计和实现方法.     

一种用于 UWB 雷达实时信号处理的可重构多处理机结构

超宽带（UWB）雷达信号处理是个多任务、大运算量的过程，必须采用多处理机并行处理技术。本文对UWB雷达实时信号处理系统的特点进行了深入的分析，在比较各种多处理机系统优缺点的基础上，提出了一种可重构的UWB雷达实时信号处理系统结构。将此结构应用于实际系统表明，该结构具有良好的性能，并具有可重构性和扩展性。     

一类多维双重时序 AR(1)-MA(q) 模型的平稳性

讨论了一类多维双重时序ＡＲ（１）－ＭＡ（ｑ）模型的二阶平稳性问题。通过导出一组线性方程组，给出了这类模型二阶平稳的显式充分条件和必要条件，为验证模型的二阶平稳性提供了一个可行的途径。     

环形区域极点/协方差约束下的模型简化设计

提出离散随机系统模型简化的一种新方法,即环形区域极点／协方差约束下的模型简化方法。这种设计方法的基本思路是构造指定维数的降阶模型,使其匹配给定的环形区域极点和稳态协方差参数,在系统的动态特性和稳态特性方面逼近给定的满阶模型。文中导出了期望的简化模型的存在条件及解析表达式,并提供了一个数值算例。     

C~3I系统的决策时延分析

决策是C~3I 系统的核心问题之一。用Petri 网理论研究C~3I 系统的决策问题已取得了长足的进展。根据决策过程各阶段之所需时间具有随机性及不独立性的特点,以变迁发射时间不独立的随机Petri 网作为决策时延的建模与分析工具,并给出了这种随机Petri 网模型的一般分析方法。     

On lifetimes in random environments

For a component operating in random environment, whose hazard rate is assumed to be the realization of a suitable increasing stochastic process, conditions are found such that its lifetime is increasing in likelihood ratio (ILR). For the lifetimes of two components of the same kind some comparisons based on partial stochastic orders are presented. Some applications to the case of repairable components are finally provided. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 365–375, 1998     

On gamma estimation via matrix kriging

In financial engineering, sensitivities of derivative prices (also known as the Greeks) are important quantities in risk management, and stochastic gradient estimation methods are used to estimate them given the market parameters. In practice, the surface (function) of the Greeks with respect to the underlying parameters is much more desired, because it can be used in real‐time risk management. In this paper, we consider derivatives with multiple underlying assets, and propose three stochastic kriging‐based methods, the element‐by‐element, the importance mapping, and the Cholesky decomposition, to fit the surface of the gamma matrix that can fulfill the time constraint and the precision requirement in real‐time risk management. Numerical experiments are provided to illustrate the effectiveness of the proposed methods.     

Analysis of artillery shoot‐and‐scoot tactics

Firing multiple artillery rounds from the same location has two main benefits: a high rate of fire at the enemy and improved accuracy as the shooter's aim adjusts to previous rounds. However, firing repeatedly from the same location carries significant risk that the enemy will detect the artillery's location. Therefore, the shooter may periodically move locations to avoid counter‐battery fire. This maneuver is known as the shoot‐and‐scoot tactic. This article analyzes the shoot‐and‐scoot tactic for a time‐critical mission using Markov models. We compute optimal move policies and develop heuristics for more complex and realistic settings. Spending a reasonable amount of time firing multiple shots from the same location is often preferable to moving immediately after firing an initial salvo. Moving frequently reduces risk to the artillery, but also limits the artillery's ability to inflict damage on the enemy.     

基于Kullback-Leibler距离离散度的加权代理模型

复杂系统的仿真通常具有高维度、高计算量等特点,代理模型因其明晰的数学表达和良好的计算特性可用于逼近真实系统。加权模型对比单个代理模型来说,其稳定性和适应性更广。不同的代理模型其性能不一,根据特定指标,可以构造最优加权代理模型。基于代理模型预测分布以及Kullback-Leibler距离构造各子代理模型之间的离散度,并提出一种新的权函数构造方法。算例表明,该方法与最优子模型的精度相当,同时能提高对真实响应分布的逼近。