面向模型粒度控制的多分辨率建模方法

Model Granularity Control Oriented Multi-Resolution Modeling

多分辨率建模仿真技术已经广泛应用在各个领域,目前的多分辨率建模仿真方法不能有效、系统、全面地解决4个关键问题:分辨率控制、模型一致性、信息损失和跨分辨率交互。针对复杂系统多分辨率建模的4个问题,提出了一种面向模型粒度控制的多分辨率建模方法。该建模方法的核心是2个建模组件:分辨率状态图和分辨率连接桥,前者利用Unified Modeling Language(UML)状态图控制由模型交互、人机交互和资源优化等触发的模型粒度变化,后者对多分辨率模型之间的关系进行独立全面的解耦建模,实现了基于信息差的解聚聚合以及一致性维护。利用系统变阶问题验证面向模型粒度控制的多分辨率建模本文的建模方法。

Multi-resolution modeling and simulation has been widely applied in various fields, but the current multi-resolution modeling and simulation methods cannot effectively, systematically and comprehensively address four key issues: resolution control, model consistency, information loss and cross resolution interaction. Addressing these four issues, Model Granularity Control Oriented Multi-Resolution Modeling was proposed. This modeling approach has two core modeling components: resolution state chart and resolution connecting bridge, the former is to use UML state diagram to control granularity changes triggered by cross-resolution interaction, human-computer interaction and resource optimization model, and the latter is to model the relationship between different resolution models independently and thoroughly, and to realize consistency maintenance based on information difference. The system variable order problem was used to validate the modeling method.