摘要
根据知识不完备的复杂系统中部分变量的定量信息,合理地构造系统相关变量间的半定量约束,是建立复杂系统定性定量相结合模型的有效方法。为了实现这一目标,一种新的系统半定量关联矩阵的构建方法被提出,它以系统部分变量观测数据流的定性划分为基础,应用灰色GM(1,1)模型建立其系统动态包络,并以动态包络中变量各种状态的对应面积之比作为它们的关联系数。这种关联矩阵不仅客观地反映了变量关联性的强弱,提高了对变量关联性的表示能力,也为灰色定性仿真过程中定量信息传播提供了一种可行的策略。
It is an effect method to model complex system with qualitative-quantitative information to correctly build the semi-quantitative restrict between variables of system according to quantitative information of part variables in incomplete knowledge complex system. A new method to build the semi-quantitative system relating matrix was advanced, it was constructed via the dynamic envelopes, which were built by GM （1, 1） based on the qualitative binning of observing data current of part variables, and its elements were calculated by the ratio of variable states＇ corresponding areas in the dynamic envelopes. This relating matrix not only truly reflects the relationship among variables, improves its expressing ability for the relationship among variables, but also provides a feasible strategy to promulgate quantitative information in grey qualitative simulation （GQSIM） process.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2007年第3期474-477,共4页
Journal of System Simulation
基金
暨南大学珠海学院引进人才科研启动基金项目(04JZKY006)
珠海市科技计划项目(PB2006101)。
作者简介
黄元亮（1963-）,男，湖南永顺人，土家族，博士，副教授，研究方向为定性仿真、灰色理论和人工智能。
