毕业设计（论文）-基于多辨率S变换的电能质量暂态扰动分类方法大论文.docx

PAGE PAGE 30 毕业设计论文基于多分辨率S变换的电能质量暂态扰动分类方法 学 生 姓 名： 班 级：电自1317 学 号：2013316021724 指 导 教 师： 所 在 单 位： 电气工程学院 答 辩 日 期： 年 月 日东北电力大学本科毕业设计论文摘 要随着复杂敏感的新型电力负荷的迅速发展，电力用户对电能质量的要求不断提高，电能质量问题受到越来越广泛的关注。本文针对电能质量的短时扰动的分类问题，提出了一种基于多分辨率S变换和模糊模式识别的短时电能质量的分类方法。首先通过多分辨率S变换将暂态扰动进行变换和重构。变换的结果为一个二维复数矩阵，对矩阵各元素求模后得到S模矩阵。S变换产生的矩阵含有丰富的扰动特征信息。为特征值的提取提供数据基础。然后再根据扰动的特征性及常用的特征量，按照，少特征量，少冗余的原则，进行特征量提取。然后采用基于语言值的电能质量分类，选取形状为平顶高斯型的隶属函数。该方法直接用待分类的各个类别来定义模糊输入变量的语言值,根据各个类别对应的输入变量取值来设定隶属函数。最后，为提高电能质量暂态扰动模糊分类效率，提出一种基于类别语言值的电能质量扰动分类方法。该方法针对电能质量扰动模糊分类器简化模糊规则需求与传统模糊推理机制之间的差异，将电能质量暂态扰动的类别作为隶属函数的语言值，给出基于类别语言值的模糊分类器设计方法，并确定隶属函数，生成模糊规则，实现分类目的与语言值选择及隶属函数设计的融合，为分类器模糊推理过程及模糊规则库的精简提供了重要依据。关键词：暂态电能质量扰动，S变换，模糊模式识别，扰动分类ABSTRACT With the rapid development of complex and sensitive new power load, the demand for power quality of electric power users is increasing, and the power quality problem is paid more and more attention. In this paper, a classification method of short-term power quality based on multi-resolution S-transform and fuzzy pattern recognition is proposed for the classification of short-term disturbance of power quality. First, the transient perturbation is transformed and reconstructed by multiresolution S transform. The result of the transformation is a two-dimensional complex matrix, and the S-matrix is obtained by modulating the elements of the matrix. The matrix generated by the S transform has rich perturbation feature information. Provide the data base for the extraction of eigenvalues. And then according to the characteristics of the disturbance and the commonly used feature quantity, according to the principle of less feature, less redundant principle, the feature extraction. Then, based on the classification of power quality based on language value. This method directly defines the language value of the fuzzy input variable with each category to be classified, and sets the membership function according to the input variable value corresponding to each category.Finally, in