摘要
目前国内外对MOA(金属氧化物避雷器)在线运行状态的研究大多是基于单脉冲下开展的,而真实自然闪电大多是多脉冲形式。单脉冲在能量上和持续时间上均远小于真实自然闪电,并不能模拟实际自然环境中的多脉冲闪电情况。为此,在多脉冲冲击试验条件下得到MOA老化劣化变化过程中的真实数据,通过分析不同类型数据的变化来研究表征MOA老化劣化的参数指标,并通过神经网络算法对数据进行分析。结果显示,提出的基于多脉冲下的神经网络算法对不同运行状态的MOA正确识别率超过96%,表明其具有高精度和正确性。为多脉冲下MOA老化劣化的研究提供完整的智能化在线监测新方法。
At present,the research of MOA( metal oxide arrester) on line is mostly based on single pulse,and the real natural lightning is mostly multi pulse. The single pulse is far less than the real natural lightning in the energy and duration,and can not simulate the multi pulse lightning in the actual environment.Therefore,in this paper,the real data in the process of aging deterioration of MOA were obtained under the condition of multi pulse impact test,and the parameters of MOA aging deterioration were studied by analyzing the changes of different types of data. The results show that the proposed neural network algorithm based on multiple pulses has a correct recognition rate of MOA over 96% in different running states,which shows that it has high-precision and accuracy. The neural network algorithm is used to analyze the data,which provides a new method for the study of aging degradation of MOA under multi pulse.
作者
王梧熠
杨仲江
于忠江
孙涌
姚旭麟
李洪阳
WANG Wuyi;YANG Zhongjiang;YU Zhongjiang;SUN Yong;YAO Xulin;LI Hongyang(Key Laboratory for Aerosol-Cloud-Precipitation of China Meteorological Administration,Nanjing University of Information Science and Technology,Nanjing 210044,China;The School of Atmospheric Physics,Nanjing University of Information Science and Technology,Nanjing 210044,China;Beijing lightning protection facilities Testing Service Center,Beijing 100089,China)
出处
《电瓷避雷器》
CAS
北大核心
2019年第1期31-37,共7页
Insulators and Surge Arresters
基金
国家重点基础研究发展计划(973计划)资助项目(编号:2014CB441405)
国家自然科学基金资助项目(编号:41075025)
江苏高校优势学科建设工程资助项目(PADA)
关键词
MOA
多脉冲
在线监测
神经网络
MOA
multi-pulse
on-line monitoring
neural network