A novel HCBR(hierarchical case based reasoning) method for auto-body assembly sequence planning was presented. Firstly, assembly information was modeled, and assembly planning was defined as a case. Through hierarchical representation and organization of the assembly information knowledge, an index scheme of case base Bayesian network was established. A computation method of comprehensive similarity was implemented to retrieve the most similar case for reuse. After case matching and adaptation, the feasible assembly sequences of aim case were finally obtained. The effectiveness of the method was verified by the results of front rail assembly and comparison

of related algorithms.

assembly sequence planning, case based reasoning, automobile body assembly modeling, similarity

针对车身装配顺序规划，提出基于分层实例推理(hierarchical case based reasoning,HCBR)方法。首先，建立车身装配信息模型，定义装配顺序规划的实例;然后,通过装配信息知识的分层表示与组织，建立实例库贝叶斯网络的索引结构，提出综合相似度算法检索最相似的重用实例;最后,通过实例匹配和修改,得到目标实例的可行装配顺序。车身前纵梁部件装配实例验证结果和算法对比表明了该方法的有效性。



装配顺序规划, 基于实例推理, 车身装配建模, 相似度