【作者】 王海柱；

【导师】 蔡泽祥；

【作者基本信息】 华南理工大学 ， 电力系统及其自动化， 2014， 博士

【摘要】 作为智能电网的关键节点，智能变电站以信息共享为根本理念，以网络化传输作为实现手段，以信息流作为基本载体，既承担着保护、计量、控制等重要业务的实现，又是构筑全网统一数据平台的基石，成为智能电网发展与创新的理想突破口，近年来取得了很大发展。然而，目前智能变电站分析方法和手段停留在定性的层面，具体表现为规划与设计依靠流量估算或者设计人员的经验判断、运行与维护依赖第三方的网络分析仪或者实测、可靠性的保证依赖于设备无约束的冗余，缺乏科学严密的推导、分析和论证，根本原因在于信息流定量分析和优化控制方法及手段的缺失。本文以智能变电站信息流为研究对象，针对信息流的定量分析、优化控制方法及手段问题展开研究。针对智能变电站信息流稳态的定量分析方法缺乏问题，本文首次构建了智能变电站信息流的潮流模型和计算方法，并结合报文广播域的划分建立了能够反映节点间逻辑连接关系的逻辑网络模型，然后从各信息源发包行为的不同出发，建立了节点模型。进一步地，给出了信息流潮流模型，并将信息流潮流计算分为稳态流量、极限流量、特定应用场景流量以及网络异常流量4类，满足了对通信网络不同运行方式进行分析的需求。最后给出了典型智能变电站信息流潮流计算的算例，并通过与网络分析仪实测的结果进行比对，验证了所提模型和方法的有效性。本文进一步提出了基于OPNET(Optimized Network Engineering Tools)的智能变电站信息流动态性能建模与仿真方法，分析了IEC61850标准、IED(Intelligent ElectronicDevice)、交换机、通信网络、仿真场景等关键对象的建模要求，利用OPNET模型库和自定义建模方法，建立智能变电站信息流动态性能仿真平台，以典型变电站为对象，分析了排队策略、交换路径、流控策略等关键因素对信息流动态性能的影响，为开展智能变电站信息流动态性能分析提供了一个有效的研究工具。针对智能变电站信息流的交换过程的不确定问题，本文首次提出了“交换延时确定、交换路径确定、交换流量确定”的确定性交换技术。研究了信息流的不确定交换问题，提出确定性交换技术，研究了基于SCD(Substation Configuration Description)解析的信息流识别与标记、基于时间戳的信息流交换延时测量、基于APPID(Application Identification)的信息流交换路径、基于层次化调度的信息流交换流量控制四大核心技术。进一步，与交换机厂家合作研发了具有电力系统业务感知能力的智能交换机，为破解智能变电站信息流的实时性可靠性问题提供可靠的技术手段。基于以上成果，本文首次提出了具备确定性交换、“网采网跳”、共网传输、不依赖外部时钟同步、电力系统业务感知五个主要特征的智能变电站通信网络优化方案，以某220kV典型变电站为实例，以信息流分析为基本方法，分析了优化方案的稳态和动态性能，构建了不依赖于外部时钟的采样值同步实验系统，验证本文方案的可行性和优越性。本文工作得到国家自然科学基金项目（51377026）、广东省自然科学基金（S2013010012885）、广东省科技计划项目（2010B01090028）以及南方电网公司科技项目（K-GD2011-505）的资助，部分研究成果已经在电网运行部门的实际工作中得到应用，验证了本文工作的正确性和有效性，取得了良好的社会和经济效益。更多还原

【Abstract】 A smart substation, an important part of Smart Grid, has developed rapidly in the lastfew years and been realized to be a breakthrough of smart grid developing, which has thefundamental idea of information sharing, and has the network as the implementation ofinformation transmission and has the traffic flow as the basic carrier of information. However,the analysis method remains on the stage of qualitative, which mainly shows up in the factthat the planning and design mainly rely on traffic flow estimation or the experience ofdesigners and the operation and maintance rely on network analysis equipment, and thenetwork reliability relies on the redundant without restriction. The fundamental reason lies inthe lack of the quantitative analysis and optimal control method of the secondary system. Inthis paper, the traffic flow of the smart substation is studied, as well as the quantitativeanalysis method and optimization control of traffic flow.Aiming at the problem of lack of quantitative analysis method for steady-state trafficflow in smart substation, this paper presents an information flow model for process layernetwork of smart substation for the first time. Firstly the physical network model includingthe topology information of the process layer is built, and with combination of the packetsbroadcast domain divided in switches, the logical network model is established to reflect thelogical connection relationship. This paper then, through analyzing the sending packetsbehavior of each information sources, builds the node model. Furthermore, the informationflow model is given, which is divided into steady-state traffic, limit traffic, specific scenariostraffic, and network anomaly traffic to meet different need of the analysis of differentoperation modes. Finally, it gives two typical examples, and uses network analyzer to verifythe validity of the models and methods in this paper.A method of dynamic simulation of the traffic flow in a smart substation is developedbased on the Optimized Network Engineering Tools (OPNET) modeler for dynamic-statequantitative analysis. The modeling need of IEC61850, Intelligent Electronic Devices (IED),switches and the simulation scenes is analyzed, and the models are constructed using theOPNET model library and custom modeling. A typical substation is taken as an example to study the impact of queuing policy, transmission path, traffic management, etc. on thedynamic performance of traffic flow, which verifies that the proposed method is an effectivetool to do research on the performance of traffic flow.Aiming at the problem of traffic flow uncertain switching, the concept of determinedswitching delay, determined transmission path and determined traffic load is proposed for thefirst time in this paper. Uncertainties of the traffic flow in the network are also analyzed. Andthe four core technologies, including the identification and marking of traffic flow based onthe analysis of the Substation Configuration Description (SCD) documents, the measurementof switching delay based on the time stamp, the chosen of the transmission path based on theApplication Identification (APPID) and the traffic management based on the hierarchicallyscheduling, are studied. Based on these technologies, an intelligent switch withbusiness-aware is developed cooperated with the PTSwtich Power technology co., LTD inGuangzhou, which can meet the performance requirement of traffic flow.Based on the results above, a prioritization scheme with the characteristics includingdetermined switching, network sampling and tripping, all traffic flow in one network,synchronization without external clock and business-aware is proposed for the first time.Finally, a typical220kV substation is taken as an example to analyze the static and dynamicnetwork performance of the prioritization scheme, which shows the superiority and feasibilityof the prioritization scheme.This research is supported by the National Nature Foundation of China (51377026), theGuangdong Nature Foundation of China (S2013010012885), Guangdong Province Scienceand technology program project (2010B01090028) and the science and technology project ofChina Southern Power Grid (K-GD2011-505). Part of the research achievement has beenapplied in operation, which verifies the effectiveness of the work in this paper.更多还原