摘要： 针对标准的QSIM算法对于复杂系统推理求解时会产生不可控分枝问题，提出一个分解仿真方法---SRSIM算法。应用聚类分析方法将定性模型分解成单元，对单元分开单独应用QSIM推理，对单元间的关系通过约束传播使得子模型推理一致。这种分解仿真算法可大大降低QSIM算法的不可控分枝，提高系统的仿真效率。
关键词:
QSIM算法,
模型分解,
定性推理
Abstract: To aim at the intractable branching problem, which the standard QSIM algorithm generates when the system is complex, the paper proposes the spliting and simulation algorithm - SRSIM. SRSIM applies the methods of cluster analysis to devide the qualitative model into some components, separtely simulates with QSIM for the components, and makes the consistence among the components through the constraint propagation. The algorithm may greatly decrease the intractable branch and improve the simulation efficiency.
Key words:
QSIM algorithm,
Model decomposing,
Qualitative reasoni
中图分类号:
杨善林, 梁昌勇, 黄梯云. 一种基于单元分解仿真算法-SRSIM[J]. 计算机工程, 2000, 26(12): 7-8+48.
YANG Shanlin, LIANG Changyong, HUANG Tiyun. The Simulation Algorithm Based on Component Decomposing-SRSIM[J]. Computer Engineering, 2000, 26(12): 7-8+48.