文献类型：期刊文献

中文题名：电力分散控制系统的过载数据自动检测算法

英文题名：Overload Data Automatic Detection Algorithm for Power Distributed Control System

作者：朱珠

第一作者：朱珠

机构：[1]贵州理工学院电气工程学院

第一机构：贵州理工学院电气与信息工程学院

年份：2015

卷号：36

期号：5

起止页码：607-611

中文期刊名：电力与能源

外文期刊名：Power & Energy

语种：中文

中文关键词：电力系统;控制;数据检测;经验模态分解

外文关键词：power system; control; data detection; empirical mode decomposition

摘要：电力分散控制系统是通过对电力网络数据库中的数据信息进行管理和调度实现电力控制的嵌入式系统,对电力分散控制系统中的过载数据自动化检测是保证电力分散控制系统稳定运行的关键。传统方法采用神经网络控制方法实现过载数据自动检测,由于神经网络的嵌入式神经元的自组织性和非线性特点,导致检测过程中信噪比过低,检测精度不高。提出一种基于经验模态分解和累积量切片后置聚焦的电力分散控制系统的过载数据自动检测算法。构建电力分散控制系统的数据结构模型,对控制系统中的数据特征进行信号模型构建,采用FIR级联滤波器对控制系统的干扰信息进行合理滤除,进一步对过载数据进行经验模态分解,通过累积量切片后置聚焦实现对信号的时频特征优化提取,达到数据自动检测的目的。仿真实验进行了性能验证,展示了该算法在提高电力分散自动控制系统中的过载数据准确检测中的优越性能,过载数据准确检测概率达到100%,优于传统算法。
Power distributed control system is embedded system for achieving power control through managing and scheduling the power network database information,and the overload data automatic detection in the distributed control system is the key to guarantee stable operation of power distributed control system.The traditional methods adopt neural network control method to realize overload data automatic detection,but the selforganizing and nonlinear characteristics of the embedded neurons in the neural network results in excessively low signal-to-noise ratio in the detection process and relatively low detection accuracy.This paper proposes an overload data automatic detection algorithm based on empirical mode decomposition and cumulant slice rear focusing as follows：construct the data structure model and the signal model based on the distributed control system data characteristics;adopt FIR cascade filters to filter the interference information in a reasonable manner;conduct the empirical mode decomposition of the overload data;optimally extract the time-frequency characteristics of signals through cumulant slice rear focusing to achieve the automatic data detection.The simulation experiment has verified the superior performance of this algorithm in improving the overload data automatic detection accuracy in the distributed control system,with the detection accuracy probability up to 100%,better than the traditional algorithm.