基于资源特征融合的复杂型面配合决策方法  

Decision-making method for complex surface coordination based on resource feature fusion

作　　者：阴艳超[1] 蒋维 徐凯[1] 廖伟智 YIN Yan-chao;JIANG Wei;XU Kai;LIAO Wei-zhi(Faculty of Mechanical and Electrical Engineering,Kunming University of Science and Technology,Kunming 650500,China;School of Mechanical and Electrical Engineering,University of Electronic Science and Technology,Chengdu 610054,China)

机构地区：[1]昆明理工大学机电工程学院,昆明650500 [2]电子科技大学机械与电气工程学院,成都610054

出　　处：《控制与决策》2021年第6期1368-1376,共9页Control and Decision

基　　金：国家重点研发计划课题(2017YFB1400301)。

摘　　要：针对复杂产品虚拟装配过程中零部件间匹配关系描述不清晰、反映装配约束的特征属性考虑不足、装配序列规划中零件配对精度不高等问题,提出基于资源特征融合的复杂型面配合决策方法.在分析典型复杂型面结构体的装配工艺关联、匹配型面特征,以及装配设计意图的基础上,构建包括工程语义(engineering semantic,ES)、单元件(single element,SE)、装配约束(assembly constraints,AC)、装配端口(assembly port,AP)、装配空间关系(assembly space relation,ASR)等多特征融合的特征装配集,采用有向无环图对特征装配集进行模块化封装和拓扑排序,获取最优装配序列,并通过模糊理想解法完成复杂型面结构体特征装配集模型从定性输入到定量求解,再到定性输出的映射变换,实现特征装配集间的精准匹配,给出匹配决策过程.最后,通过应用案例验证模型与算法的有效性,为实现复杂型面结构体高效精确装配提供新思路.Aiming at the problems of unclear description of matching relationship between components in virtual assembly process of complex products, the narrow consideration of characteristic attributes reflecting technological constraints,and the low accuracy of parts matching, a decision-making method for complex surface coordination based on resource feature fusion is proposed. Firstly, by analyzing the assembly process association, matching surface features, and assembly design intent of a typical complex surface structural body, a feature assembly set based on multi-feature fusion including engineering semantics(ES), single element(SE), assembly constraints(AC), assembly ports(AP), and assembly space relationships(ASR) is constructed. Then, the assembly set is modularly encapsulated and topologically sorted using directed acyclic graph. Through using the fuzzy ideal solution, the transition process that the assembly set model from qualitative input to qualitative solving and then to qualitative output can be realized, and the matching decision process is given. Finally, an application case is given to illustrate the effectiveness of the proposed model and the algorithm, which provides a new idea for efficient and accurate assembly of complex surface structures.

关 键 词：虚拟装配 特征装配集 多资源特征 模块化封装 拓扑排序 匹配决策 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]