中文摘要：

选择拆卸序列规划是产品维修或回收的重要环节，针对目前选择拆卸序列规划算法中自动化程度较低的问题，提出一种基于运动规划的选择拆卸序列规划方法。该方法首先根据复杂产品中零件数量繁多，形状不规则的特点，采用基于自适应动态多树的快速扩展随机树（Rapidly-exploring random tree，RRT）算法对零件进行运动规划。在此基础之上，通过对装配体进行自动分层处理，分析零件间拆卸约束关系，构建装配体的拆卸约束关系图。最后通过对拆卸约束关系图的分析处理，获得目标零件的选择拆卸序列。以某底盘的目标零件为例，对提出的算法进行了验证。

英文摘要：

Selective disassembly sequence planning is one of the most important stages in product repairing and recycling. Due to the lack of generalization of existing methods, a motion planning based selective disassembly sequence planning method is proposed. Aiming at the feature of huge amount and irregular shapes of complex product parts, a motion planning method based on dynamic self-adapted multiple trees based rapidly-exploring random tree （RRT） algorithm is introduced to plan the disassembly path of the parts initially. Based on that, a hierarchical structure of the assembly is generated which leads to the generation of the disassembly constraint relationship graph of all the assembly parts. By analysis of the generated graph, the disassembly sequence of selected part can be obtained. Results from test on a part from a complex chassis assembly validate this approach.