基于业务风险的智能电网通信端到端QoS保障及评估模型研究

Smart Grid relies greatly on the design, development, and deployment of real-time, reliable and secure data communication networks that enable information communication between devices, applications, consumers and grid operators. Timing is critical in Smart Grid communications, which is the most fundamental difference from other communication networks. Some types of information exchanges between electric devices are useful only within a predefined time frame. If the communication delay exceeds the required time window, the information does not serve its purpose anymore and, in the worst case, damage might be incurred in the grid. The communication infrastructure in Smart Grid will incorporate many network technologies and assume a hierarchical and hybrid composition. The messages communicated between various entities within the power grid, may have different network latency requirements. Therefore, how to guarantee the end-to-end QoS for communications in Smart Grid is a great challenge. In this project, 1)the risk and QoS metrics are studied, and a Model of Dynamic Business Priorities Related with Context is provided; 2)the components of end-to-end delay are analyzed and evaluation methods of delays for those components are researched; 3)based on business risk analysis, the architectures of communication networks in Smart Grid to support end-to-end QoS guarantees with context-aware QoS control techniques and managment policies are proposed; 4)suggestions for improving protocols and QoS control techniques aim at communications in Smart Grid are given. We pointed that they are trade-offs between meeting the QoS requirements and risks for Grid to guarentee QoS for communications, and limiting end-to-end delay must avoid congestion and link-faults in netowrks. We consider they are possible to design error-recoverable protocols and to control end-to-end QoS crossing layer. This work can direct the design and managment of communication networks in Smart Grid.

实时、可靠、安全的通信网络是智能电网实现的基础。本项目针对智能电网通信中的实时业务对端到端时延要求苛刻、通信网络结构和技术复杂、应用类型多样的特点，研究智能电网通信网络的端到端QoS保障体系结构及评估模型，包括：1)业务风险和QoS指标体系，环境相关的业务动态优先级模型；2)业务的端到端时延模型；3)基于业务风险和QoS的网络结构与资源分配优化算法，支持端到端QoS保障的网络体系结构，环境感知QoS控制和管理机制；4)网络协议和QoS算法或机制针对智能电网通信特点的改进。项目提出了基于业务风险建立智能电网通信网络的QoS保障体系结构的思想和环境相关业务动态优先级的概念；提出端到端时延的评估要综合考虑网络正常运行的固有时延和异常时拥塞和链路恢复等引入的时延；提出将跨层优化方法引入端到端QoS控制机制设计和将纠错机制引入网络协议设计。成果将对智能电网通信网络建设和管理、标准制定具有指导意义。

实时、可靠、安全的通信网络是智能电网实现的基础。随着大区电网互联对广域系统保护需求的提出，电力控制业务的实时性保障已成为智能电网进一步发展亟待解决的问题。为此，本项目开展了如下主要工作：1）针对智能电网通信网络技术和结构不同于公用互联网、协议专有的问题，研究了智能电网通信网络的QoS分析方法，建立了仿真与理论分析相结合的基本策略；研究了在主流仿真软件中对智能电网通信网络进行建模的方法，建立了实际的仿真模型；分析了不同架构和技术的实时性，得出了相应的结论；建立了智能电网通信业务的端到端时延模型和QoS保障体系结构；2）分析了智能电网信息空间存在的脆弱性，研究了网络攻击对电力调度数据网性能的影响；考虑对电力一次系统的影响，研究了智能电网信息空间的脆弱性评估方法、和考虑网络攻击因素与交互影响的业务风险评估指标体系和方法；研究了电力骨干通信网资源的优化问题，提出了基于业务风险的路由优化分配遗传算法模型；3）针对现有智能变电站两级业务优先级的不足，提出了新的优先级划分方法和调度算法；针对使用TCP传输不能保证业务实时性和UDP传输不能保证可靠性的问题，提出了一种新的广域保护系统业务实时性与可靠性保障方案，并针对电力业务报文特点设计了相应算法；针对现有基于数字证书的公钥密码体制不能保证实时性的问题，提出了一种新的满足智能变电站业务实时性要求的通信安全认证体系，并设计了相应的协议改进方法；4）针对网络安全问题越来越突出，研究了智能电网安全态势感知方法与智能电网中易受攻击网络通信的安全保障技术。得到如下结论：通信业务的实时性不仅受到通信网络自身技术和结构的影响，同时受到网络安全状态的影响，在当前敌对势力高级持续性攻击的威胁下，对实时业务的QoS保障需要重点考虑网络攻击因素；智能电网已发展成为将计算、通信、控制与物理系统深度融合的多维复杂系统，研究方法必须站在信息物理融合系统的整体视角上，并考虑各子系统之间的交互影响。成果将对智能电网通信网络建设和管理、标准制定具有参考意义。