电力系统稳态数据参数化压缩算法  

Parametric Compression Algorithm for Power System Steady Data

作　　者：张斌[1] 张东来[1] 

机构地区：[1]哈尔滨工业大学深圳研究生院,广东省深圳市518055

出　　处：《中国电机工程学报》2011年第1期72-79,共8页Proceedings of the CSEE

基　　金：国家自然科学基金项目(50977016)~~

摘　　要：针对电力系统运行中产生的大量稳态数据,提出一种基于自适应神经模糊推理系统(adaptive-network-based fuzzy inference system,ANFIS)建模的参数化压缩算法。通过实现等相位采集方式,有效改善电网稳态数据的周期滑步现象,并建立和分析等相位数据的数学模型,为参数化压缩算法提供理论依据。设计参数化压缩算法的结构,并给出实现流程。该算法利用ANFIS系统建立基准序列模型,用于重构数据,采用三次样条插值建立基准相位模型,用于计算各周波相位差,由此将需存储的数据变为模型参数及相位差。最后利用仿真和实测数据详细分析影响系统信噪比和压缩比的因素,为参数的选取提供了参考,验证了算法不仅能提高稳态数据的压缩比,而且能有效滤除系统噪声,为电力系统海量数据的压缩提供了一种新思路。A parametric compression algorithm based on adaptive-network-based fuzzy inference system（ANFIS） model was proposed for the vast amounts of steady data in power system.The equiphase acquisition devices improve the steady-state data cycle sliding step phenomenon effectively.A mathematical model of equiphase data sampling was established to provide conditions for the parametric compression algorithm.The algorithm structure was designed and detailed process was described.ANFIS is employed to create the data model in the algorithm,which is used to reconstruct sampling data,and cubic spline interpolation is employed to establish reference phase model and then to calculate phase difference of each cycle.Finally,simulating and experimental data show the impacts of different factors on compression ratio（CR） and signal noise ratio（SNR） in order to provide references for the parameters selected.It proves that this algorithm can greatly improve the compression ratio and de-noising effect.This paper demonstrates a new approach for data compression in power system.

关 键 词：电力系统 稳态数据 参数化压缩 自适应神经模糊推理系统 等相位采样 

分 类 号：TM71[电气工程—电力系统及自动化]