中文摘要：

通过优化网络拓扑结构,提出了一种在家庭能源管理系统（HEMS）中基于动态规划的排队论聚类算法.首先将HEMS的总花费划分成网关的安装成本及数据传输成本2个部分,然后通过对比2种不同的排队论聚类算法,将家庭能源管理系统中的成本问题转化为系统网关部署问题,最后,设计了一种使系统花费最小的M2M网关配置方案.仿真结果表明,当网关缓冲区间足够大时,HEMS系统的花费主要来自网关的安装成本.当网关缓冲区间有限时,通过边际分析方法及成本最小化特性,所提出的排队算法能够保证在最小化花费的基础上实现网关的最优分配.

英文摘要：

By optimizing the network topology, this paper proposes a newmethod of queuing theory clustering algorithm based on dynamic programming in a home energy management system（ HEMS）. First, the total cost of the HEMS system is divided into two parts, the gateway installation cost and the data transmission cost. Secondly, through comparing two kinds of different queuing theories, the cost problem of the HEMS is converted into the problem of gateway deployment. Finally, a machine-to-machine（ M2M） gateway configuration scheme is designed to minimize the cost of the system. Simulation results showthat the cost of the HEMS system mainly comes from the installation cost of the gateways when the gateway buffer space is large enough. If the gateway buffer space is limited, the proposed queue algorithm can effectively achieve optimal gateway setting while maintaining the minimal cost of the HEMS at desired levels through marginal analyses and the properties of cost minimization.