基于PCA和EEMD的柔性直流配电网故障选线算法

Fault Line Selection Algorithm for Flexible DC Distribution Network Based on PCA and EEMD

柔性直流故障选线技术的发展对直流配电网有着至关重要的作用。本文针对现有柔性直流故障存在的可利用的故障信息较少等问题,提出了一种新算法,该算法有效利用了集合经验模态分解EEMD(ensemble empirical mode decomposition)算法、主成分分析PCA(principal component analysis)和相关系数各自的优势。首先,提取暂态电流样本信号,采用EEMD得到以正交基函数表示的数据矩阵;接着,基于PCA进行该矩阵元素特征向量到主成分的转换,将样本信号投影到主元空间实现坐标变换,从而得到对样本数据的聚类和识别结果;最后,基于相关系数进行故障线路判别。本文算法的EEMD揭露了原始历史数据的内在变化规律,PCA能够有效选择故障有效特征。大量实验表明,该新算法准确有效,与现有其他方法相比,在故障信息不明显、不同过渡电阻方面具有优势。

The development of the flexible DC fault line selection technology plays an important role for DC distribution network.In this paper,a novel algorithm is proposed to solve the problem that there is less available fault information about the existing flexible DC fault,which makes full use of the advantages of ensemble empirical mode decomposition(EEMD),principal component analysis(PCA)and the correlation coefficient algorithm.First,the transient current sample signal is extracted,and the data matrix represented by the orthogonal basis function is obtained by EEMD.Then,the feature vector of the matrix element is transformed into the principal component based on PCA,and the sample signal is projected into the principal component space to realize coordinate transformation,so as to obtain the clustering and identification results of the sample data.Finally,fault line identification is performed based on the correlation coefficient.The EEMD of the proposed algorithm reveals the internal variation law of the original historical data,while PCA can effectively select the effective fault features.A large number of experiments show that the novel algorithm is accurate and effective.Compared with other existing methods,it has advantages in the cases of unclear fault information and different transition resistances.