摘要
模糊神经网络既具有处理非线性问题的优势,同时又具备模糊的特性,使其在智能信息处理中具备强大的先天优势。本研究首先介绍了模糊控制基础理论,分析了基于RNN的控制技术和CNN的控制技术,然后结合基于深度学习提出LSTM长短期记忆单元的模糊控制技术,并与传统控制技术相比较,并经过仿真实验筛选出基于长短期记忆单元的控制决策技术,应用于电气系统故障诊断。
Fuzzy neural network not only has the advantage of dealing with nonlinear problems,but also has the characteristic of fuzzy,so it has a strong innate advantage in intelligent information processing.Firstly introduces the fuzzy control theory,this study analyzes the control technology based on RNN and CNN's control technology,and then combined with based on the deep learning LSTM both short-term and long-term memory unit of the fuzzy control technology,and compared with the traditional control technology,and through simulation experiment based on short-and long-term memory unit selection control decision-making technology,applied to electrical system fault diagnosis.
作者
邹昊东
王会羽
孙琦
张彬彬
ZOU Hao-dong;WANG Hui-yu;SUN Qi;ZHANG Bin-bin(Information&Communication Branch/State Grid Jiangsu Electric Power Company,Nanjing 210024,China;Anhui Jiyuan Software Co.,Ltd.,Hefei 230088,China)
出处
《山东农业大学学报（自然科学版）》
北大核心
2020年第4期708-711,共4页
Journal of Shandong Agricultural University：Natural Science Edition
基金
国网江苏省电力有限公司科技项目:基于数字孪生模型的机房运行环境监测及一站式调控的关键技术研究与应用(J2018022)。
关键词
模糊神经网络
智能电器系统
故障诊断
Fuzzy neural network
intellectual electric system
fault diagnosis
作者简介
邹昊东(1987-),男,硕士,工程师,主要从事数据中心运维,云计算技术研究和应用等工作.E-mail:puzhengguo@sgitg.sgcc.com.cn;通讯作者:张彬彬.E-mail:857446812@qq.com。
