基于数字孪生自动色环机的设计知识图谱研究  

Research on Knowledge Graph of Automatic Coloring Machine Design Based on Digital Twin

作　　者：王佳皓 张太华[2] WANG Jia-hao;ZHANG Tai-hua(Key Laboratory of Modern Manufacturing Technology of Ministry of Education,Guizhou University,Guiyang 550025,China;School of Mechanical and Electrical Engineering,Guizhou Normal University,Guiyang 550025,China)

机构地区：[1]贵州大学现代制造技术教育部重点实验室,贵阳550025 [2]贵州师范大学机械与电气工程学院,贵阳550025

出　　处：《组合机床与自动化加工技术》2022年第4期15-19,共5页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金(72061006)。

摘　　要：结合数字孪生和知识图谱技术,对自动色环机设计知识图谱进行研究。首先,进行了基于数字孪生的虚拟设计与本体建模;其次,建立了基于数字孪生和知识图谱的自动色环机设计框架,包含初本体建模、知识抽取、知识融合与知识推理;最后,结合强化算法和蚁群算法,建立了知识推理算法框架模型,将蚁群算法中的信息素作为策略网络进行实体间的路径选择,并对蚁群算法中的信息素更新进行了改进与比较,最终训练得到了预期结果。In this paper, the design knowledge graph of automatic color ring machine is studied by combining digital twin and knowledge graph technology.Firstly, virtual design and ontology modeling based on digital twin are carried out.Secondly, a design framework of automatic color ring machine based on digital twin and knowledge graph was established, including initial ontology modeling, knowledge extraction, knowledge fusion and knowledge reasoning.Finally, the framework model of knowledge inference algorithm was established by combining the reinforcement algorithm and ant colony algorithm.The pheromone in ant colony algorithm was used as the strategy network to select the path strength between entities.The pheromone update in ant colony algorithm was improved and compared.

关 键 词：数字孪生 知识图谱 知识推理 强化学习 蚁群算法 

分 类 号：TH166[机械工程—机械制造及自动化] TG502[金属学及工艺—金属切削加工及机床]