《粗糙集故障诊断中的应用研究》-毕业设计(论文).doc

精品论文 精品论文 粗糙集故障诊断中的应用研究 摘 要 电力变压器作为电力系统中最重要的设备之一，其故障诊断技术的研究一直是国内外学者关注的热点。粗糙集不需要先验的知识，可以直接从数据中挖掘有用的信息；神经网络具有自学习能力，经常用于分类模型的建立。本文将粗糙集理论和神经网络两者相结合，应用于变压器故障诊断中，构建粗糙集和神经网络智能混合系统。首先利用粗糙集理论对知识的约简能力提取出对于故障诊断有主要影响的因素。然后将神经网络用于约简后的数据训练样本，建立变压器故障诊断分类模型。此基础构建的神经网络规模大大减小，学习速度大为提高，而又保持了网络较好的分类能力。最后结合变压器历史故障样本数据进行仿真，分别建立了BP和RBF神经网络故障诊断模型，结果表明本文的方法是有效的。 关键词： 电力变压器 粗糙集 BP神经网络 RBF神经网络 故障诊断 On Fault Diagnosis Based on Rough Sets Theory Abstract Because the transformer is one of the most important equipments in power system, the technology of fault diagnosis for transformer is always taken into account by savants all over the world. Rough sets does not require a priori knowledge, it can digging useful information from the data directly; The neural network has self-learning ability, it often be used in the establish of classification model. The theory combines rough set theory and neural networks and applied to transformer fault diagnosis to build intelligent hybrid system of rough sets and neural networks. First, the rough set theory for knowledge reduction ability to extract the main factors for fault diagnosis. Then the neural network training sample after data reduction to establish the transformer fault diagnosis classification model. It has greatly reduced network scale and greatly improved learning speed, while keeps the good classification ability of network. Finally, combine the historical fault sample data to simulate, then establish BP and RBF neural network fault diagnosis model respectively, the result show that this method is effective. Key Words: Transformer; Rough Set; BP Neural Network; RBF Neural Network; Fault Diagnosis 目 录 TOC \o 1-3 \h \z \u 摘 要 I Abstract II 第一章 绪论 1 1.1课题研究的背景和意义 1 1.2 课题的任务及解决方法 2 1.3 论文的结构 3 第二章 课题相关的理论基础 4 2.1 粗糙集理论 4 2.1.1 粗糙集理论的概况 4 2.1.2 粗糙集方法的特点 5 2.1.3 粗糙集理论的基础知识 5 2.2 连续属性离散化 7 2.2.1 离散化的概念 7 2.2.2 常用的离散化策略 7 2.3 人工神经网络 9 2.3.1人工神经网络的结构 9 2.3.2 BP神经网络 10 2.3.3 RBF神经网络 14 第三章 变压器故障