摘要
变压器是电力系统中最重要的枢纽设备之一,其保护动作的正确是整个电力系统稳定运行的关键。对于接入大规模风电场的系统,双馈风电机组在低电压穿越期间短路电流特殊的频偏特性将导致传统变压器差动保护方案在风电场送出变中动作性能变差。针对这一问题,首先推导了频偏特性下短路电流经DFT提取后的误差表达式,并分析该短路电流对传统保护的影响。在此基础上,提出了利用BP神经网络模型来逼近变压器电磁暂态模型,并结合波形相关性进行故障识别的新方案。最后,在MATLAB/SIMULINK平台上搭建双馈风电场仿真模型。仿真结果表明,在各运行工况下,基于BP神经网络模型的方案能够有效进行故障识别,并规避频偏特性及励磁涌流带来的影响。
The power transformer is one of the most important equipments in a power system.Its reliable protection system is the key to the stable operation of the whole power system.The unique frequency offset characteristics of short-circuit current of doubly-fed wind farm(DFIG)during low-voltage ride through will affect the differential protection performance in the transmission transformers.To solve this problem,firstly,the error expression of short-circuit current extracted by DFT under frequency offset characteristics is introduced,and the influence of short-circuit current on traditional protection is analyzed.On this basis,this paper proposed a new scheme using BP(Back Propagation)neural network model,combined with waveform correlation for fault identification,to approximate the transformer electromagnetic transient model. Finally, the simulation model of a doubly fed wind farm is built on MATLAB/SIMULINK platform to verify this scheme.The results show that it can effectively identify faults under various operating conditions, and minimize the impact of frequency offset characteristics and inrush current.
作者
杨兴雄
孙士云
黄柯昊
YANG Xingxiong;SUN Shiyun;HUANG Kehao(Faculty of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650500,China)
出处
《电力科学与工程》
2021年第11期23-31,共9页
Electric Power Science and Engineering
基金
国家自然科学基金重点项目(52037003)
云南省重大专项(202002AF080001)。
关键词
双馈感应风力发电机
频偏特性
变压器故障识别
BP神经网络
波形相关性
doubly-fed induction generator
frequency offset characteristics
transformer fault identification
BP neural network
waveform correlation
