摘要
新型电力系统中广泛应用电力电子装置,导致谐波成分复杂且随时间变化。有效检测这些谐波分量对于提升电能质量与效率具有重要意义。集合经验模态分解能够求解电力系统信号的频谱信息,并提取各次谐波的幅值、相位等特征但需人为设定参数。提出了一种基于深度网络的神经网络训练自适应谐波分解模型,该模型能够自动选择参数,无须依赖人工设置,从而避免了人为因素引入的误差。最后,验证了所提出方法的有效性。
The new power systems contain a large number of power electronic devices,resulting in complex har-monic content,with harmonic frequencies varying over time.Monitoring the harmonic components can improve the power quality and efficiency of these systems.Empirical Mode Decomposition(EMD)can resolve the spectral information of power system signals,such as the amplitude and phase of each harmonic.This paper proposes an adaptive harmonic decomposition model based on deep neural networks,where the model parameters can be auto-matically selected without relying on manual settings,thus avoiding human-induced errors.Finally,test cases are presented to verify the effectiveness of the proposed method.
作者
姚兵
申冉
YAO Bing;SHEN Ran(DC Company,State Grid Hubei Electric Power Co.,Ltd.,Yichang 443000,Hubei Province,China;Yichang Power Supply Company,State Grid Hubei Electric Power Co.,Ltd.,Yichang 443099,Hubei Province,China)
出处
《电力与能源》
2024年第6期660-663,共4页
Power & Energy
关键词
深度网络
神经网络
新型电力系统
谐波分离
deep network
neural network
new power system
harmonic separation
