基于数据驱动的时序概率谐波电流预测研究  

Research on data-driven sequential probabilistic harmonic current prediction

作　　者：丁玉昊 刘建波 曹立明 谢香敏 DING Yuhao;LIU Jianbo;CAO Liming;XIE Xiangmin(College of Electrical Engineering,Qingdao University,Qingdao 266071,China;State Grid Dongying Kenli District Power Supply Company,Dongying 257500,China)

机构地区：[1]青岛大学电气工程学院,山东青岛266071 [2]国网东营市垦利区供电公司,山东东营257500

出　　处：《供用电》2023年第10期19-27,45,共10页Distribution & Utilization

基　　金：山东省自然科学基金项目(ZR2021QE056);中国博士后科学基金面上项目(2022M721759)。

摘　　要：随着非线性电力电子装置的广泛应用,配电网中的谐波含量日益增加,为有效地应对谐波日益严重的负面影响,需对谐波电流进行预测。居民负荷是常见的谐波源,由于其具有很强的随机性,导致其产生的谐波电流数据具有概率性波动。为此,提出一种基于RBF-ARX模型的时序概率谐波电流预测方法。首先,采用RBF-ARX模型对谐波电流的确定性分量进行点预测;然后,采用统计学的概率预测方法预测谐波电流的随机性分量;最后,基于实测谐波数据,验证了所提时序概率谐波电流预测方法的准确性与实用性。With the widespread use of nonlinear power electronic devices,the harmonic content caused by them is increasing.To effectively counter the increasingly negative impact of harmonics,it is necessary to predict the harmonic currents.As a common harmonic source,residential loads are highly randomized,resulting in probabilistic fluctuations in their harmonic current data.For this reason,this paper proposes a time series probabilistic harmonic current prediction method based on RBF-ARX model.Firstly,the RBF-ARX model is used to predict the deterministic component of harmonic current in time series;then the probabilistic prediction method of statistics is used to predict the stochastic component of harmonic current.Finally,the accuracy and practicability of the time-series probabilistic harmonic current prediction method proposed in this paper is verified based on the measured harmonic data.

关 键 词：RBF-ARX模型 深度神经网络 高斯混合分布 概率谐波电流预测 马尔可夫蒙特卡洛采样法 

分 类 号：TM74[电气工程—电力系统及自动化]