摘要
为了解决自动售烟机机械结构模块化设计时,采用模糊聚类算法得出的模块划分结果无法得到有效验证的问题,提出一种根据运动学和动力学仿真数据来证模糊聚类结果是否合理的方法。首先,根据各个零件的功能、几何、物理相关性建立模糊聚类原始矩阵,通过夹角余弦公式将原始矩阵转换为模糊相似矩阵;然后,由模糊相似矩阵通过传递闭包的算法得到模糊等阶矩阵,由模糊等阶矩阵取不同的阈值形成动态聚类树;最后,由Adams软件对模糊聚类后的核心模块进行运动学和动力学仿真,由仿真结果判断模糊聚类的结果是否合理。实验表明,使用所提出方法进行模糊聚类结果判定时,可以有效判断模糊聚类的结果是否合理。
In order to solve the problem that the partition results obtained by using fuzzy clustering algorithm cannot be verified effectively in the modular design of the mechanical structure of cigarette vending machine,this paper proposes a method to verify the rationality of fuzzy clustering results based on kinematics and dynamics simulation data.Firstly,the original fuzzy clustering matrix is established according to the function,geometry and physical correlation of each part,and the original matrix is transformed into the fuzzy similarity matrix by the included angle cosine formula.Then,the fuzzy equal-order matrices is obtained by using the transitive closure algorithm,and the dynamic clustering tree is formed by taking different thresholds from the fuzzy equal-order matrices.Finally,Adams software is used to simulate the kinematics and dynamics of the core module after fuzzy clustering,and the simulation results are used to judge whether the fuzzy clustering results are reasonable or not.The experiment shows that the fuzzy clustering results can be judged effectively by using the method presented in this paper.
作者
谢任林
张志强
Xie Renlin;Zhang Zhiqiang(Beijing Information Science&Technology University,Beijing 100192,China)
出处
《电子测量技术》
北大核心
2021年第7期19-25,共7页
Electronic Measurement Technology
基金
北京市教委科研计划科技一般项目(KM202011232011)资助
关键词
自动售烟机
模块化设计
模糊聚类
运动学仿真
automatic cigarette vending machine
modular design
fuzzy clustering
dynamic simulation
