一种支撑协同制造的动态群组命令传输算法  

Dynamic group command transfer algorithm for collaborative manufacturing

作　　者：颜晓莲 章刚 邱晓红 Yan Xiaolian;Zhang Gang;Qiu Xiaohong(School of Software Engineering,Jiangxi University of Science&Technology,Nanchang 330013,China;Peking University Science Park,Nanchang 330013,China)

机构地区：[1]江西理工大学软件工程学院,南昌330013 [2]北大科技园,南昌330013

出　　处：《计算机应用研究》2020年第8期2362-2365,2394,共5页Application Research of Computers

基　　金：江西省教育厅科技项目(GJJ170571)。

摘　　要：Internet是工业物联网重要网络载体之一,由于其时刻动态变化,使得在支撑群组控制命令传输时会造成路径失效过期,导致无法对分散在不同区域的生产设备实现协同制造。首先把Internet群组控制命令传输问题转换为带约束的动态多Steiner树问题(constrainted dynamic multi-Steiner trees,CDMST),并提出一种动态遗传算法。该算法提出静态和动态两个维度混合搜索策略,当网络状态未发生变化时,执行静态搜索策略,静态搜索策略主要考虑缓解早熟现象。当网络状态发生变化时,执行动态搜索策略,动态搜索策略主要考虑提升收敛性及借鉴静态搜索策略缓解早熟现象。实验从群组命令传输成功率和收敛延时比验证该算法有效性、收敛性及全局搜索能力。Internet is one of the important network carriers of the industrial Internet of Things.Because of its dynamic changes,the path will expire when supporting the transfer of group control commands,which cannot achieves collaborative manufacturing to equipments scattered in different regions.This paper transformed the group control command transfer problem into constrained dynamic multi-Steiner trees problem firstly,and proposed a kind of improved dynamic genetic algorithm.The algorithm proposed a hybrid search strategy with static and dynamic dimensions.It executed static search strategy when network state didn’t change,static search strategy mainly considered alleviating premature phenomena.It executed dynamic search strategy when network state changed,dynamic search strategy mainly considered improving convergence and using static search strategy to alleviate premature phenomena.This paper verified the validity,convergence and global search ability of the algorithm from transfer success rate,convergence delay rate.

关 键 词：工业物联网 群组命令传输 协同制造 动态Steiner树 动态遗传算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]