因果定性推理是一种通过分析描述物理系统行为和关系的约束找出系统内部各个成分之间的因果结构的推理方法．本文提出一种基于约束和变量分析的因果定性分析模型和算法．该方法在产品设计中有广泛的应用，利用这个模型和算法可较好地解决参数化设计中的几何推理问题，还可用作概念设计的工具，用于完成复杂系统设计任务的划分及定序、设计变量之间相互依赖关系分析等工作．算法具有应用性强、效率和稳定性好、支持欠约束和多解问题等优点.

Causal qualitative reasoning is aimed to explore the causalities betweencomDonents in a system by analyzing the descriptions for physical behaviors and relation-shiDs of the system．This paper presents a model and algorithms of causal qualitative reasoning based 0n constraint and variable analysis．The result can be widely applied to many aspects of product design systems，such as parametric design by geometric reason-ing，decomposition of complex design tasks and dependency analysis of design parameters.The algorithms have advantages in applicability，efficiency，robustness and ability to deal with 11nder-constrained conditions in a constraint system．

本文研究得到国家自然科学基金资助.