摘要
在光伏高比例接入的孤岛微电网中,太阳辐照度变化将导致光伏出力明显波动,继而影响微电网线路故障电流变化趋势。对于不同辐照度光伏出力,如何准确完成微电网故障辨识是微电网故障保护研究面临的一个重要问题。针对这一问题,首先建立了含有光伏发电的典型微网模型,分析了孤岛模式下太阳辐照度变化对网内故障电流的影响;然后利用快速小波能量熵(wavelet energy entropy,WEE)算法提取线路故障电流的暂态特征,继而选取简洁清晰的暂态特征并构建故障综合样本集;最后,利用典型故障样本集对极限学习机进行训练,形成一种新的考虑光伏出力间歇性的微网故障辨识方法。算例仿真结果表明所提方法既能精确提取不同辐照度下微网故障的暂态特征,又可准确地实现故障辨识,为微电网故障分析和保护提供了技术支持。
In the islanded microgrid integrated with high proportion of photovoltaic power,the change of solar irradiance will lead to obvious photovoltaic power fluctuation,and then affect the trend of fault current of lines in microgrid.For the photovoltaic power with different irradiance,how to accurately identify the microgrid fault is important for microgrid protection.To this issue,a typical microgrid model with photovoltaic power generation is firstly established in this paper,and the influence of solar irradiance variation on the fault current in the islanded microgrid mode is analyzed.Then,the fast wavelet energy entropy(WEE) algorithm is used to extract the transient characteristics of fault current,and the concise transient characteristics are selected to construct the fault comprehensive sample set.Finally,considering photovoltaic power intermittence,this paper forms a new microgrid fault identification method by training Kernel-based extreme learning machine with typical fault sample set.The simulation result shows that the proposed method can not only accurately extract the transient characteristics of microgrid fault under different irradiance,but also accurately identify the fault,which provides technical support for the fault analysis and protection of microgrid.
作者
李浩茹
丁保迪
季宇
王永刚
陈继开
张利伟
LI Haoru;DING Baodi;JI Yu;WANG Yonggang;CHEN Jikai;ZHANG Liwei(School of Electrical Engineering,Northeast Electric Power University,Jilin 132012;Jilin Province,China 2.China Electric Power Research Institute,Beijing 100085)
出处
《电力建设》
北大核心
2020年第9期76-85,共10页
Electric Power Construction
基金
国家电网公司总部科技项目“基于多能互补的微网规划与运行控制技术研究及应用”(52060018000N)。
关键词
微电网
光伏发电
故障辨识
特征提取
快速小波能量熵(WEE)算法
microgrid
photovoltaic power generation
fault identification
feature extraction
fast wavelet energy entropy(WEE)algorithm
