一个典型C~3I系统的微分方程作战模型的建立

本文针对目前C~3I系统建模研究遇到的难题:如何把C~3I系统的组成内容与它所处的战争环境结合起来,运用复杂系统的一些建模方法、基于问题导向的原则,提出了一套战争环境下C~3I系统建模的规范化建模方法,使C~3I系统与军方常用于描述作战的Lanchester型格斗方程有机地结合起来,形成C~3I系统的作战方程。文中以一个典型的防空作战的C~3I想定进行了具体的计算,建立了该想定的微分方程模型。结果表明,这种方法解决了C~3I系统与Lanchester方程的结合问题,从而使C~3I系统本身的各个组成部分能在模型中与作战环境有机地结合起来。     

多专家集成式决策支持系统探讨

本文在文献[1]给出集成式决策支持系统(I—DSS)的基本概念及提出I—DSS的两种主要形式的基础上,对I—DSS的形式之一——多专家集成式决策支持系统(MEI—DSSS)作了较深入的探讨。本文阐述了MEI—DSS提出的原因,给出了MEI—DSS的基本概念,概述了MEI—DSS同相关学科和技术的关系,提出了MEI—DSS的基本结构框架,并从知识表示、推理机制以及多专家系统动态组合进行问题求解的算法实现等方面探讨了该系统在大型决策中的应用。本文最后提出了MEI—DSS今后有待深入研究的课题。     

Decision Support Systems (DSS) development using a model of user inquiry types: Methodological proposals and a case study

A study of the DSS implementation area reveals an increased emphasis on methodologies for DSS development. As the DSS field matures, a larger number of methodological options are becoming available to the DSS developer. Existing methodologies have adopted various methods for performing the important task of DSS requirements analysis and specification. In envisaging a hierarchy of requirements analysis methods, a method for improved requirements analysis is proposed to remedy observed deficiencies within it. This method is based on a view of generic types of inquiries or solicitations made by the DSS user during decision-making. This formalism constitutes the basis of a proposed DSS development methodology that offers several benefits to DSS development. The intended benefits include a sharper and more focused requirements analysis to improve the DSS development process. This paper describes the model and methodology, alongside details of a real-world example of their use in developing a Marketing Decision Support System (MKDSS) for the support of marketing decision-making.     

非线性系统的模糊辨识在线算法

在以往的模糊辨识方法中,均未给出在线调整非线性系统的模糊辨识算法。基于递推模糊聚类方法,提出了一种在线调节模糊模型的辨识算法。为了验证所提出方法的有效性,对几个非线性系统进行了辨识,最后给出了辨识结果。     

混合过程控制系统基于广义对象的形式化开发

许多过程控制系统可以看作混合控制系统。为使设计的控制器满足整个系统一定的过程需求,对象的信息是不可缺少的。本文为这类系统的开发提出了广义对象的形式化模型。基于这样的模型,可以从目标要求中直接提取系统需求,控制器也可以根据基于对象的形式化方法进行设计。我们以一个工业过程控制的实例来说明我们的方法和模型,采用时段演算来描述系统的特性并进行控制器的开发。     

广义不确定分布参数系统变结构控制

研究了一类广义不确定分布参数系统的综合问题,由广义不确定分布参数系统的特点,直接给出了变结构控制的设计方法。设计过程简单,控制易于实现,且给出了实例     

基于变神经网络的非线性最小方差预测控制器

提出基于变神经网络学习动态系统参数的最小方差预测控制器。其目的是通过在线学习 ,使控制器(MVPC)能适应被控对象参数变化和非确定性。提出的变神经网络由两部分组成 ,一部分是线性神经网络 (LNN) ,作为被控对象局部线性动态模型 ,另一部分是多层交叉回归神经网络 (DRNN) ,它近似为非线性动态模型。由于引进递推最小方差算法 ,本控制器运算速度相当快。仿真结果表明所提方法对非线性系统自适应控制是有效的     

区间Lurie时滞系统鲁棒绝对稳定性的LMI方法

对于区间Lurie时滞控制系统 ,构造了关于Lurie型Lyapunov泛函中正定矩阵和积分项系数的线性矩阵不等式 (LMI) ,通过LMI的解构造的Lyapunov泛函来保证系统的鲁棒绝对稳定性 ,不必选择和调整参数。分别讨论了无穷扇形角和有限扇形角的情形 ,所获得的结果适用于系统具有多个非线性执行机构的情形。最后给出一个实例说明本方法的有效性 ,并通过实例分析了扇形角的大小与鲁棒稳定度的关系。     

Qualifying Function in SSM Modeling—A Case Study

In a previous paper the author introduced a concept called qualifying function as a conceptual tool for modeling and design. The concept was incorporated into the modeling phase of soft systems methodology (SSM), with the purpose of addressing some of the criticism directed toward the methodology, such as its tendency to result in regulatory, rather than radical, agendas for change. It was argued that the concept had the potential to help people in a particular problem situation to view the situation from new and different perspectives and that it could enrich the conceptual models by giving the transformation process a direction. The aim of this paper is to apply the concept, embedded in SSM, to a case study involving two projects, focused on creating new work opportunities, to explore the practical applicability and usefulness of the concept, both in general and as a part of the modeling phase of SSM.     

Action Research: Its Role in the University/Business Relationship

The Action Research (AR) PhD program at Monash University had its genesis in Systems Thinking and Organizational Learning. This paper discusses the role of the university in AR projects in business and the central role that AR projects can have in a Faculty's strategic positioning. The issues that have emerged in the project to date are discussed.