The MGM(1,m,N)model is an effective grey multi-variate forecasting model that considers multiple system characteristic sequences affected by multiplefactors.Nevertheless,it isregularly inaccurate in the application Th...The MGM(1,m,N)model is an effective grey multi-variate forecasting model that considers multiple system characteristic sequences affected by multiplefactors.Nevertheless,it isregularly inaccurate in the application This is because the model requires a strong correlation between the system characteristic sequences That reduces the applicability of the model.To solve this problem,this paper proposes a novel multi-variate grey model.This model does not require a certain correlation between system characteristic sequences and has higher applicability Through numerical integration,a two-point trapezoidal formula,and a recursive method,thetime-response expressions ofthetwo model forms are obtained Some properties of the proposed model are further discussed Finally,the validity of the proposed model is evaluated by using two real cases related to China's invention patent development.Theresultsshow that the novel models outperformother models inbothsimulation and prediction applications.展开更多
配电网状态监测系统建设是构建泛在电力物联网的重要内容,准确预测通信网络带宽对明确业务需求、优化资源配置具有重要意义。针对当前配用电基于排队论的通信带宽预测方法在混合业务并存时带宽计算过程复杂的局限性,提出了一种基于G/M/...配电网状态监测系统建设是构建泛在电力物联网的重要内容,准确预测通信网络带宽对明确业务需求、优化资源配置具有重要意义。针对当前配用电基于排队论的通信带宽预测方法在混合业务并存时带宽计算过程复杂的局限性,提出了一种基于G/M/1/N队列模型的通信带宽预测方法。从状态监测系统架构和数据业务特点出发,分析混合数据业务到达时间间隔的统计描述以及基本传输速率,并给出了满足服务质量(quality of service,QoS)要求的系统带宽最优化化求解方法。以典型配电网状态监测应用为例,讨论了业务带宽、服务质量和带宽利用率之间的量化评价过程;验证了配电网状态监测网络信息流的"小数据"特征,为低功耗广域物联网通信技术在配电网状态监测中的应用提供了理论支撑。展开更多
基金Supported bythe National Natural Science Foundation of China(71701105)the Major Program of the National Social Science Fund of China(17ZDA092)+1 种基金the Key Research Project of Philosophy and Social Sciences in Universities of Jiangsu Province(2018SJZDI111)Key Projects of Open Topics of Jiangsu Productivity Society in2020(JSSCL2020A004)。
文摘The MGM(1,m,N)model is an effective grey multi-variate forecasting model that considers multiple system characteristic sequences affected by multiplefactors.Nevertheless,it isregularly inaccurate in the application This is because the model requires a strong correlation between the system characteristic sequences That reduces the applicability of the model.To solve this problem,this paper proposes a novel multi-variate grey model.This model does not require a certain correlation between system characteristic sequences and has higher applicability Through numerical integration,a two-point trapezoidal formula,and a recursive method,thetime-response expressions ofthetwo model forms are obtained Some properties of the proposed model are further discussed Finally,the validity of the proposed model is evaluated by using two real cases related to China's invention patent development.Theresultsshow that the novel models outperformother models inbothsimulation and prediction applications.
文摘配电网状态监测系统建设是构建泛在电力物联网的重要内容,准确预测通信网络带宽对明确业务需求、优化资源配置具有重要意义。针对当前配用电基于排队论的通信带宽预测方法在混合业务并存时带宽计算过程复杂的局限性,提出了一种基于G/M/1/N队列模型的通信带宽预测方法。从状态监测系统架构和数据业务特点出发,分析混合数据业务到达时间间隔的统计描述以及基本传输速率,并给出了满足服务质量(quality of service,QoS)要求的系统带宽最优化化求解方法。以典型配电网状态监测应用为例,讨论了业务带宽、服务质量和带宽利用率之间的量化评价过程;验证了配电网状态监测网络信息流的"小数据"特征,为低功耗广域物联网通信技术在配电网状态监测中的应用提供了理论支撑。