基于总体最小二乘改进的SDFT三相交流电频率估计算法

发布时间：2018-06-07 22:07

本文选题：频率估计 + 总体最小二乘　；参考：《东南大学学报(自然科学版)》2017年06期

【摘要】：为了解决SDFT算法在含有噪声、谐波或突发干扰的电力网络中电压信号基频分量三点关系式不严格成立的问题,引入总体最小二乘算法(TLS-SDFT)进行改进.TLS-SDFT算法采用滑动窗截取的多个DFT基频分量样本点来扩展SDFT算法的三点关系式,引入扰动矩阵,并通过对系数矩阵进行奇异值分解,使扰动矩阵具有最小Frobenious范数,得到改进的频率估计值.由于系数矩阵结构的特殊性,该算法的额外复杂度为窗长的线性函数.仿真结果表明,在高斯白噪声干扰下,改进算法的估计偏差和均方误差远低于原SDFT算法.在高次谐波干扰、信号参数突变以及变电站实测环境下,改进算法的频率追踪精确度均有明显提升.
[Abstract]:In order to solve the problem that the fundamental frequency component of the voltage signal in a power network with noise, harmonic or burst interference is not strictly established in the SDFT algorithm, the total least square algorithm (TLS-SDFT) is introduced to the improved.TLS-SDFT algorithm to extend the three point formula of the SDFT algorithm by using the multiple DFT basic frequency component sample points of the sliding window. The perturbation matrix is introduced and the singular value decomposition of the coefficient matrix is carried out to make the perturbation matrix have the minimum Frobenious norm, and the improved frequency estimation is obtained. Due to the particularity of the coefficient matrix structure, the additional complexity of the algorithm is a linear function of the window length. The simulation results show that the improved algorithm is estimated under the Gauss white noise interference. The error and mean square error are far lower than the original SDFT algorithm. In the high order harmonic interference, the signal parameter mutation and the substation measurement environment, the improved algorithm has a significant increase in frequency tracking accuracy.
【作者单位】：东南大学信息科学与工程学院;
【基金】：国家自然科学基金资助项目(61401094,61771124) 江苏省自然科学基金资助项目(BK20140645) 中央高校基本科研业务费专项资金资助项目(2242016K41050)
【分类号】：TM711;TM935

【相似文献】