基于贝叶斯网络时序模拟推理的配电网可靠性评

【中文摘要】 配电系统可靠性评估是电力系统可靠性评估中重要环节之一。配电系统作为电力系统中电网直接与广大电力客户相连的部分,其可靠性是供电企业和电力客户非常关心的问题,因而也是非常重要的研究方向。配电网可靠性的研究,其目的是向电力客户提供安全可靠的电力供应,从而获得最优的经济效益和最佳的社会效益。配电网可靠性评估的算法是进行配电网可靠性研究的重要课题。在常规可靠性计算中,仅计算系统各项可靠性指标,很少涉及可靠性最薄弱环节的识别,而识别电力系统可靠性最薄弱环节,具有非常重要的意义。本文在综述了一些常用配电网评估算法基础上,提出了一种基于贝叶斯时序网络的配电系统可靠性评估模型。将贝叶斯网络时序模拟推理算法应用于配电系统可靠性评估中,利用前向采样技术,利用贝叶斯网络对不确定性知识的灵活表示,利用“用户配电系数”,较好地进行了配电网的可靠性指标评估,较好的解决了配电系统灵活的因果推理和诊断推理,准确的识别了钳制配电系统可靠性的的薄弱环节。系统的模拟基于基本元件的模拟,利用元件加速稳定法,能较快的加速元件模拟推理和系统模拟推理的收敛速度。以Microsoft Visual C#为开发工具开发了配电系统可靠性评估软件。该软件运用高级面向对象技术和图形化的手段,运用本文提出的基于贝叶斯网络时序模拟推理算法对配电系统进行可靠性指标的计算,并定量地分析配电系统的薄弱环节,通过对数据的分析处理,得出较传统方法更有意义的结论。最后,用IEEE-RBTS bus2作为实验系统验证了此方法的正确性和软件的有效性。因此,基于贝叶斯网络时序模拟推理算法的配电系统可靠性评估是一种较好的方法。
【英文摘要】 The distribution systems reliability assessment is one of the most important parts of power systems reliability assessment. Because distribution system is the direct joining part between power network and the vast power customers, its reliability is the issue power supply enterprise and customers are very concerned about and also a very important research direction. Distribution networks reliability research’s purpose is to provide customers with safe and reliable power supply, thus obtains the best economic and social benefits. The algorithm of distribution systems reliability assessment is an important issue on the study of distribution systems reliability. Reliability conventional calculation involes various system reliability indexs, but rarely touches upon the identification on the weakest links of reliability. However, the identification has a very important significance. This paper presents a kind of distribution systems reliability assessment model based on sequence bayesian networks after summarizing some commen distribution systems reliability assessment algorithms. This method build a bayesian network model of distribution system reliability assessment firetly, apply sequence bayesian networks simulation reasoning algorithm to distribution system reliability assessment, make use of forward facing sampling technique, make use of bayesian networks’flexible representation about uncertain knowledge, make use of "User Distribution Coefficient", carry out a better assessment of reliability indicators of distribution network, make a solution of the distribution system’s flexible causal reasoning and diagnostic reasoning and identify accuratly the weak links of clamp distribution system reliability.The system simulation is based on the simulation of basic components. The compenents and system simulation reasoning’s convergence rate can be accelerated using accelerated stable methods of components.A kind of distribution system reliability assessment software is developed with Microsoft Visual C# as a development tool. The software make a calculation about reliability indicators of distribution network using sequence bayesian networks simulation reasoning algorithm with a high-level object-oriented technology and graphical means, quantitatively analyses the weak links in the distribution system and obtain more meaningful conclusions than traditional methods by data analyzing and processing. At last, IEEE-RBTS bus2 Experimental System proves the method’s correctness and software’s effectiveness. So reliability assessment of distribution systems based on sequence bayesian networks simulation reasoning is a better method.  （科教范文网http://fw.NSEAC.com编辑发布）
【中文关键词】 配电系统; 可靠性评估; 贝叶斯时序模拟推理算法; 元件加速稳定
【英文关键词】 Distribution system; Reliability assessment; Sequence bayesian networks simulation reasoning algorithm; Accelerated stable methods of component
【论文目录】
摘要 4-5
Abstract 5-6
第一章 绪论 9-14
    1.1 配电系统可靠性概念 9
    1.2 配电系统特点 9
    1.3 配电系统可靠性评估的重要意义 9-10
    1.4 配电系统可靠性研究的任务 10-11
    1.5 国内外可靠性研究的发展与现状 11-13
        1.5.1 国外配电系统可靠性评价 11-12
        1.5.2 我国配电网系统可靠性的研究以及发展现状 12-13
    1.6 本文的研究内容和主要贡献 13-14
第二章 贝叶斯网络推理 14-22