中国电机工程学报
中國電機工程學報
중국전궤공정학보
ZHONGGUO DIANJI GONGCHENG XUEBAO
2014年
7期
1159-1166
,共8页
电能质量%谐波阻抗%波动量法%柯西分布%Pearson分布
電能質量%諧波阻抗%波動量法%柯西分佈%Pearson分佈
전능질량%해파조항%파동량법%가서분포%Pearson분포
power quality%harmonic impedance%fluctuation method%Cauchy distribution%Pearson distribution
从波动量法的基本理论入手,研究波动量比值符号的概率统计特征,通过理论分析提出当谐波源受到多种随机因素影响时波动量比值的实部和虚部的概率密度服从广义柯西分布,并选取Pearson VII公式对此进行描述。然后,给出利用概率统计和非线性曲线拟合估计谐波阻抗的具体方法及步骤。进而分析影响波动量法误差的各种因素,推导理论误差表达式并提出修正误差的措施。与各类依赖判据进行数据过滤的方法相比,概率法的数据处理过程更为简便,具有较好的通用性、可靠性及抗干扰性,同时,概率法能更全面地分析波动量比值的概率统计特征,避免数据的误处理。蒙特卡洛仿真及实例应用分析验证了所提方法的有效性和实用性。
從波動量法的基本理論入手,研究波動量比值符號的概率統計特徵,通過理論分析提齣噹諧波源受到多種隨機因素影響時波動量比值的實部和虛部的概率密度服從廣義柯西分佈,併選取Pearson VII公式對此進行描述。然後,給齣利用概率統計和非線性麯線擬閤估計諧波阻抗的具體方法及步驟。進而分析影響波動量法誤差的各種因素,推導理論誤差錶達式併提齣脩正誤差的措施。與各類依賴判據進行數據過濾的方法相比,概率法的數據處理過程更為簡便,具有較好的通用性、可靠性及抗榦擾性,同時,概率法能更全麵地分析波動量比值的概率統計特徵,避免數據的誤處理。矇特卡洛倣真及實例應用分析驗證瞭所提方法的有效性和實用性。
종파동량법적기본이론입수,연구파동량비치부호적개솔통계특정,통과이론분석제출당해파원수도다충수궤인소영향시파동량비치적실부화허부적개솔밀도복종엄의가서분포,병선취Pearson VII공식대차진행묘술。연후,급출이용개솔통계화비선성곡선의합고계해파조항적구체방법급보취。진이분석영향파동량법오차적각충인소,추도이론오차표체식병제출수정오차적조시。여각류의뢰판거진행수거과려적방법상비,개솔법적수거처리과정경위간편,구유교호적통용성、가고성급항간우성,동시,개솔법능경전면지분석파동량비치적개솔통계특정,피면수거적오처리。몽특잡락방진급실례응용분석험증료소제방법적유효성화실용성。
Starting from the basic theory of fluctuation method, this paper studied probability characteristics of the ratio of harmonic voltage fluctuation to harmonic current fluctuation. Through theoretical analysis, it was proved that the probability density of both real and imaginary part of the ratio follow the generalized Cauchy distribution if the harmonic source is affected by many random factors. So the Pearson VII formula was selected to describe the above distributions. An approach and its specific steps of using probability statistics and nonlinear curve-fitting were proposed to be used for estimation of power system harmonic impedances. Furthermore, this paper analyzed the factors affecting the estimated error of fluctuation method, and then gave out a set of error formulas and measured to reduce the error. Compared with the methods depending on data filtering strategies, the probability approach is more simple and convenient in data processing. And it has more advantages such as universality, reliability and noise immunity. Meanwhile, it can reduce failure rate of data processing by considering the probability characteristics in a wide perspective. Monte Carlo simulation and field test data analysis show that the proposed approach is practical and effective.