基于广义最小二乘回归的间谐波源识别方法  

Identification of inter-harmonic sources based on generalized least squares regression

作　　者：赵振涛[1] 张大海[1] 李永生 刘洋[1] 孙德达[1] 

机构地区：[1]山东大学电气工程学院,山东济南250061 [2]山东菏泽供电公司,山东菏泽274000

出　　处：《电力系统保护与控制》2013年第15期36-40,共5页Power System Protection and Control

基　　金：国家自然科学基金项目(50707016)~~

摘　　要：间谐波阻抗等系统参数对间谐波源的识别有着重要意义。为了克服电压和电流信号中有色噪声影响,首次将广义最小二乘回归法应用于间谐波阻抗等系统参数估计。该方法通过白化滤波器将有色噪声转化成白噪声,然后对转化的数据进行参数估计,从而克服变量与变量之间相关性的影响,与传统回归估计方法相比能够减小间谐波阻抗估计误差。此外,还结合间谐波国标,引入量化指标划分供用电双方的责任大小。仿真及实例分析验证了该方法的有效性。System parameters such as the inter-harmonic impedance are crucial to the identification of inter-harmonics sources. To suppress the effects of colored noise in the voltage and current signals, a method based on generalized least squares is proposed to estimate system parameters for the first time. This method eliminates the effect of the correlation between variables by converting the colored noise to the white noise firstly, and then carrying out parameters estimation with the converted data, which reduces inter-harmonic impedance estimation error compared with the conventional regression method. In addition, this paper introduces quantitative index to quantify the responsibilities at both sides combined with the national standard. Simulation analysis and practical testing data prove the method effective and feasible.

关 键 词：间谐波源识别 阻抗估计 线性回归 广义最小二乘 责任划分 

分 类 号：TM711[电气工程—电力系统及自动化]