摘要
通过分析线路出口故障时过渡电阻对距离保护的影响,比较传统型距离保护、自适应距离保护、神经网络距离保护的优缺点,利用南水电厂到泉水电厂双端电源供电系统进行PSASP仿真研究分析。对仿真线路进行大量试验以获得相关数据信息,计算出倾斜角。把仿真线路测得的数据结合已经计算出的倾斜角一起输入到三层BP网络进行训练,以获得一个训练好的三层BP网络。当线路发生短路时,通过把线路的实时测量数据输入到已训练的BP网络中,ANN就能估算出倾斜角,以实现人工神经网络在自适应距离保护的应用。证明基于人工神经网络的自适应距离保护具有更好的躲过渡电阻特性。
By analyzing the impact of the transition resistance on the distance protection when line outlet fault occurs,this paper compares the advantages and disadvantages among traditional distance protection,adaptive distance protection,and distance protection of neural networks.By the PSASP simulation of double-ended power supply system from Guangdong Shaoguan Power Grid South Hydro Power Plant to Spring Hydro Power Plant,it tests a large number of circuits in order to calculate the angle.The measured data together with the value of the angle are input to BP network to get a well-trained three-layer BP network.When short circuit occurs in the line,through inputting the real-time measurement data into the trained BP networks,ANN can estimate the inclination angle to realize the application of ANN in the self-adaptive distance protection.It is proven that artificial neural network based adaptive distance protection is better to escape the transition resistance.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2011年第8期124-127,154,共5页
Power System Protection and Control
关键词
短路
过渡电阻
自适应
人工神经网络
距离保护
short circuit
transition resistance
self-adaptive
ANN
distance protection