摘要
光伏系统多采用串并联(SP)连接,某一组件发生故障对整个光伏系统都将产生较大影响。如某一组件断路,则组串电流为零,整个组串功率丢失。及时发现故障并判断光伏故障类型,对光伏系统平稳运行、提升光伏发电效能具有重要作用。此处基于改进自适应共振理论(ART)神经网络,通过模拟不同组件故障,提取不同故障类型下的特征参数,进而使用ART神经网络进行训练,获得光伏阵列故障诊断模型。试验表明所提光伏故障诊断方法能够较为精准地判别短路、断路、阴影遮挡、老化等故障。
Series parallel(SP)connection is often used in photovoltaic system.The failure of a module will have a great impact on the whole photovoltaic system.If one component is open circuit,the string current is zero and the whole string power is lost.It is very important to find the fault and judge the type of photovoltaic fault in time for the smooth operation of photovoltaic system and the improvement of photovoltaic power generation efficiency.Based on the improved adaptive responance theory(ART)neural network,the characteristic parameters of different fault types are extracted by simulating different module faults,and then the ART neural network is used for training to obtain the photovoltaic array fault diagnosis model.Experiments show that the proposed photovoltaic fault diagnosis method can accurately distinguish short circuit,open circuit,shadow,aging and other faults.
作者
陈咏涛
王瑞妙
樊晓伟
朱小军
CHEN Yong-tao;WANG Rui-miao;FAN Xiao-wei;ZHU Xiao-jun(State Grid Chongqing Electric Power,Chonging 401123,China;不详)
出处
《电力电子技术》
CSCD
北大核心
2022年第8期81-84,118,共5页
Power Electronics
基金
国网重庆市电力公司科技项目(SGCQDK00DWJS2100144)。
关键词
自适应共振理论神经网络
光伏阵列
故障诊断
adaptive responance theory neural network
photovoltaic array
fault diagnosis
