基于谱分析的电力负荷特征量提取方法研究  

Study of Power Load Characteristic Quantity Extraction Based on Spectral Analysis

作　　者：赵磊[1] 仝霞[1] 李雪城 程鹏申 解进军 ZHAO Lei;TONG Xia;LI Xue-cheng;CHENG Peng-shen;XIE Jin-jun(Customer Service Center,State Grid Beijing Electric Power Company,Beijing 102600,China)

机构地区：[1]国网北京市电力公司客户服务中心,北京102600

出　　处：《机械工程与自动化》2023年第5期50-51,54,共3页Mechanical Engineering & Automation

摘　　要：针对现有的电力负荷暂态特征量提取算法基于快速傅里叶变换(FFT)而FFT算法存在频谱泄漏和栅栏效应的问题,提出了同步化-Prony谱分析算法(Syn-Prony算法)。在传统的Prony算法基础上,对采样信号进行频移预处理,使其具有采样同步性;同时为了避免时频转换过程中频谱混叠干扰问题,对同步化后的采样信号使用梳状有限冲激响应(Comb-FIR)陷波滤波器进行处理。以典型居民家庭用电负荷为对象进行了实验研究,实验结果表明,Syn-Prony方法弥补了现有负荷辨识算法在暂态特征量提取方面的不足。The existing power load transient feature extraction algorithm is based on fast Fourier transform(FFT),but the FFT algorithm has the problems of spectrum leakage and fence effect.In this paper,Syn-Prony spectral analysis algorithm(Syn-Prony algorithm)is proposed.Based on the traditional Prony algorithm,the sampling signal is preprocessed by frequency shift to make it have sampling synchronization.At the same time,in order to avoid spectrum aliasing interference in the process of time-frequency conversion,the sampled signal after synchronization is processed by Comb-FIR notch filter.An experimental study is carried out on the electricity load of typical households.The experimental results show that the Syn-Prony method makes up for the shortcomings of the existing load identification algorithms in transient characteristic quantity extraction.

关 键 词：负荷辨识 FFT 谱分析 

分 类 号：TM935.21[电气工程—电力电子与电力传动]