基于神经网络的电子式电能表消噪系统设计  

Design of Electronic Energy Meter Noise Elimination System Based on Neural Networks

作　　者：江剑峰[1] 张垠[1] 朱彬若[1] 翁素婷 

机构地区：[1]上海市电力公司电力科学研究院,上海200051

出　　处：《电器与能效管理技术》2015年第7期29-32,53,共5页Electrical & Energy Management Technology

摘　　要：为提高电子式电能表计量精度、增强通信功能的可靠性和准确性,将人工神经网络技术与数字滤波器技术相结合,针对电能表内部电子电路、通信总线典型噪声干扰,设计了消噪系统,并从理论上给出了消噪系统的工作原理与仿真实现。仿真试验结果表明,基于神经网络设计的消噪系统可以很好地消除噪声,并且不会对原始波形产生畸变影响,具有一定的实际指导意义。In order to improve the measurement accuracy of electronic electric energy meter,enhance the reliability and accuracy of communication function,this paper proposed a filter design method using the artificial neural network technique and digital filter technology.The principle of neural network was introduced.Then,aiming at the typical noise interference appeared in the electric circuit and communication bus of electronic energy meter,a noise elimination system was designed based on neural networks.Moreover,theoretical principle and simulation of the noise elimination system were presented.The simulation results show that this system can achieve good noise elimination effect and does not produce the distortion of the original waveforms.

关 键 词：电能表内部噪声 人工神经网络 数字滤波器 消噪 畸变 

分 类 号：TM933.4[电气工程—电力电子与电力传动]