摘要
针对分布式电源和不平衡负荷接入配电网致状态估计精度不足问题,提出一种基于加权高斯过程回归(WGPR)的三相不平衡配电网鲁棒状态估计方法。采用高斯过程回归模型构建量测值与状态值之间的非线性映射关系,获得状态估计值。通过神经网络状态预测模型获得状态预测值,将估计值和预测值输入WGPR模型得到最终的三相状态估计结果。以IEEE 34节点和IEEE 123节点三相不平衡配电网为例进行分析,结果表明所提状态估计方法具有较高的估计精度和鲁棒性。
Targeting the problem of low accuracy in distribution network state estimation caused by the integration of distributed power and unbalanced load,the paper proposes a weighted Gaussian process regression(WGPR)-based robust state estimation method for three-phase unbalanced power distribution network.Firstly,the Gaussian process regression(GPR)is employed to establish the mapping relationship between measurements and status value,obtaining the state estimate.Then a neural network state prediction model is applied to get the state prediction.The state estimate and state prediction are input into the WGPR model to obtain the final estimation results.Taking the IEEE 34-node and IEEE 123-node three-phase unbalanced power distribution system as an example to make an analysis,the result shows that the proposed method has good precision and robustness.
作者
吉兴全
刘小虎
张玉敏
叶平峰
王飞
任童洲
JI Xingquan;LIU Xiaohu;ZHANG Yumin;YE Pingfeng;Wang Fei;REN Tongzhou(College of Electrical Engineering and Automation,Shandong University of Science and Technology,Qingdao 266590,China;State Grid Dezhou Power Supply Company,Dezhou 253073,China;State Grid Tongchuan Power Supply Company,Tongchuan 727031,China)
出处
《智慧电力》
北大核心
2023年第11期61-68,共8页
Smart Power
基金
国家自然科学基金青年基金资助项目(52107111)
山东省自然科学基金资助项目(ZR2022ME219,ZR2021QE117)。
关键词
神经网络
高斯回归
三相不平衡
鲁棒状态估计
neural network
Gaussian regression
three-phase unbalance
robust state estimation
作者简介
吉兴全(1970),男,山东潍坊人,博士,教授,主要研究方向为配电网优化。
