电工技术2024,Issue(3) :59-62.DOI:10.19768/j.cnki.dgjs.2024.03.014

基于单片机中人工神经网的故障电弧识别研究

Research on Series Fault Arc Identification Based on Artificial Neural Network in Microcontroller

徐丽红 柯拥勤 郑少威 吴君凯 杨志 李保罡 蒋祖立
电工技术2024,Issue(3) :59-62.DOI:10.19768/j.cnki.dgjs.2024.03.014

基于单片机中人工神经网的故障电弧识别研究

Research on Series Fault Arc Identification Based on Artificial Neural Network in Microcontroller

徐丽红 1柯拥勤 1郑少威 2吴君凯 3杨志 2李保罡 2蒋祖立1
扫码查看

作者信息

  • 1. 国网福建省电力有限公司莆田供电公司,福建 莆田 351100
  • 2. 华北电力大学电子与通信工程系,河北 保定 071003
  • 3. 厦门理工学院,福建 厦门 361024
  • 折叠

摘要

提出了一种基于单片机中人工神经网络的低压串联交流故障电弧识别方法,搭建了交流串联故障电弧实验平台,设计了基于STM32F407ZGT6 单片机与 AD7606 芯片的信号采集板,建立了 TensorFlow下的人工神经网络模型.通过采集正常工作和发生故障电弧的电流奇次谐波数据,对人工神经网络模型进行训练,将训练后的人工神经网络导入单片机STM32F407ZGT6 中,用于对故障电弧进行识别.实验结果表明该方法可准确识别交流串联故障电弧.

Abstract

This paper proposed a low voltage AC series fault arc identification method based on artificial neural network in microcontroller.The experimental platform of AC series fault arc was built,the signal acquisition board based on STM32F407ZGT6 microcontroller and AD7606 chip was designed,and the artificial neural network model was constructed by using TensorFlow.The artificial neural network model was trained by odd harmonic data acquired under normal operation and fault arc induction,and subsequently input into the STM32F407ZGT6 microcontroller for fault arc i-dentification.Experiment results show that the proposed method can accurately identify AC series fault arc.

关键词

故障电弧/人工神经网络/奇次谐波/TensorFlow

Key words

fault arc/artificial neural network/odd harmonic/TensorFlow

引用本文复制引用

基金项目

国家电网福建省电力公司科技项目()

国家自然科学基金(61971190)

出版年

2024
电工技术
重庆西南信息有限公司(原科技部西南信息中心)

电工技术

影响因子:0.177
ISSN:1002-1388
参考文献量12
段落导航相关论文