智能配用电通信网关键技术研究

发布时间：2018-05-08 02:02

本文选题：配用电通信 + 资源分配　；参考：《华北电力大学(北京)》2014年博士论文

【摘要】：配用电通信网支撑电力“配电”和“用电”两个重要生产环节,是实现智能电网“双向数据互动为基础,精细化智能管理”的关键。目前,国内外缺乏对配用电通信网的系统研究,因此,需要依据配用电网络架构和业务接入节点分布规律,融合应用多种通信介质构造配用电通信网应用场景模型,研究配用电业务模型、资源分配优化模型、节点部署模型等,力求实现配用电通信网的智能拥塞管理与流量均衡,提升QoS服务质量。首先,基于配用电通信网的多介质融合应用场景和业务断面模型分析了配用电业务属性和流量特征,提出了配用电业务的静态优先级CoS参数和静态QoS参数,并基于层次分析法的多属性判断形成加权公平性QoS调度参数,建立了配用电业务模型,有利于基于业务权重动态调整和合理分配业务占用资源。其次,针对单独物理层资源分配未考虑业务中诸如延迟和丢包率等QoS要求的问题,基于资源跨层调度思想,将物理层的功率和吞吐量等指标、链路层的延迟和丢包率等指标、与应用层的业务模型等指标综合考虑,提高系统资源利用率和资源分配公平性。先提出了一种基于用户紧急度和业务公平性因子相结合的单OFDM周期内的跨层资源分配算法,在保证用户公平性的同时尽量满足各业务QOS需求；再提出一种基于QoS偏离度的混合业务公平性与速率自适应相结合的长周期(多个OFDM周期)内的跨层资源分配优化算法,满足多用户混合业务资源分配的多目标要求；再针对通信网络的资源分配包括网关间的资源分配和网关内各用户间的子载波和功率资源分配两种形式,在分析传统单网关下资源分配的基础上,提出了中继型和复合型两种多网关OFDM2A系统的分布式资源分配算法,主网关节点先将系统物理信道资源分配给子网关,进而子网关再分配给各业务用户,并通过增加子网关二级缓存和负载因子判断以实现网络负载均衡,在保证各业务QoS的前提下实现了系统吞吐量最大化。最后,针对无线传感器网络在实现配用电通信网延伸覆盖时因节点能量有限、易过早失效问题,先提出无线传感器网络在配电网应采用“链式分布为主、环式网状为辅”的弱环状的部署模型,研究改进了基于弱环状部署的AODV平面路由算法；再提出了无线传感器网络在用电网应采用非均匀蜂窝的“立体网状”的部署模型,并研究改进了基于非均匀蜂窝状分簇LEACH立体路由算法。
[Abstract]:The use of electrical communication network to support the two important production links of power distribution and electricity use is the key to realize the "two-way data interaction" and "fine intelligent management" for the smart grid. At present, there is a lack of systematic research on the distribution of electrical communication networks at home and abroad. Therefore, the distribution rules of the electric network architecture and the service access node need to be used according to the network. In order to realize the intelligent congestion management and flow balance and improve the quality of QoS service, the multi media fusion application field based on the electrical communication network is used. The view and business section model are used to analyze the property and flow characteristics of the electrical service. The static priority CoS parameters and static QoS parameters are proposed, and the weighted fairness QoS scheduling parameters are formed based on the multi attribute judgment of the analytic hierarchy process. The distribution of electrical business model is established, which is beneficial to the dynamic adjustment and rationality of the business weight. Secondly, in view of the problem that the individual physical layer resource allocation does not consider the problem of QoS requirements such as delay and packet loss rate in the business, based on the idea of resource cross layer scheduling, the index of the power and throughput of the physical layer, the delay and the packet loss rate of the link layer, and the service model of the application layer are considered and improved. System resource utilization and resource allocation fairness. First, a cross layer resource allocation algorithm based on the combination of user emergency and business fairness factor in a single OFDM cycle is proposed. In order to ensure the fairness of the user and satisfy the QOS requirements of each business as much as possible, a kind of fairness and rate of mixed services based on the deviation degree of QoS is proposed. The optimization algorithm of cross layer resource allocation in a long period (multiple OFDM cycle) is combined to meet the multi-objective requirements of multi user mixed service resource allocation, and then the resource allocation of the communication network includes two forms of resource allocation among gateways, subcarriers and power resource allocation among various users in the gateway. On the basis of resource allocation under the gateway, a distributed resource allocation algorithm for two kinds of multiple gateway OFDM2A systems is proposed. The main network joint points the system physical channel resources to the sub gateway first, and then the sub gateway is reassigned to each service user, and the network is judged by increasing the two level cache of the sub gateway and the load factor to realize the network. The network load is balanced, and the system throughput is maximized under the premise of guaranteeing the QoS of each service. Finally, the wireless sensor network should adopt the "chain distribution mainly, the ring network is supplemented" in the distribution network for the wireless sensor network, which is due to the limited energy of the node and the premature failure. The weak ring deployment model is used to improve the AODV plane routing algorithm based on the weak ring deployment, and the "three-dimensional network" deployment model of the wireless sensor network should be used in the power grid, and the research and improvement of the LEACH stereo routing algorithm based on the non-uniform honeycomb honeycomb clustering is also improved.

【学位授予单位】：华北电力大学(北京)
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM73

【相似文献】