基于协同学习的工业物联网MCS排序任务推荐研究  

Research on Industrial Things Internet MCS Ranking Task Recommendation in Iot Based on Collaborative Learning

作　　者：张紫烨 Zhang Ziye(School of Information Engineering(Software School),Henan University of Animal Husbandry and Economy,Zhengzhou Henan 450044,China)

机构地区：[1]河南牧业经济学院信息工程学院(软件学院),河南郑州450044

出　　处：《现代工业经济和信息化》2024年第10期142-143,147,共3页Modern Industrial Economy and Informationization

摘　　要：为了进一步提高物联网中移动群智感知(Mobile Crowd Sensing,MCS)参与者的积极性,结合混合模型(Hybrid model,HM)与列表级排序(List-Wise Ranking,LWR)算法的优点,设计了一种HM-LWR协同学习方法,并开展MCS排序任务推荐分析。研究结果表明:与未优化LWR相比,HM-LWR协同学习方法具备良好生长速率和理想处理性能。HM-LWR协同学习方法的参与者积极性最高,参与率达到98%左右,能够获得较高分配精度,减少任务分配所需的时间,获得更高任务执行效率,相对LWR算法具备显著优点。该研究有助于提高工业物联网的运行效果,具有很好的节能效果。In order to further improve the enthusiasm of Mobile Crowd Sensing(MC)participants in the Internet of Things,combined with the advantages of Hybrid model(HM)and List-Wise Ranking(LWR)algorithms,An HM-LWR collaborative learning method is designed and MCS ranking task recommendation analysis is carried out.The results show that compared with non-optimized LWR,HM-LWR collaborative learning method has good growth rate and ideal processing performance.HM-LWR collaborative learning method has the highest enthusiasm of participants,with a participation rate of about 98%.It can obtain high assignment accuracy,reduce the time required for task assignment,and obtain higher task execution efficiency,which has significant advantages over LWR algorithm.This research is helpful to improve the operation effect of industrial Internet of Things and has a good energy-saving effect.

关 键 词：物联网 移动群智感知 协同学习 混合模型 参与者意愿 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]