摘要
主要介绍了一种提高距离保护基频分量估计算法速度和精度的新方法 .该方法采用递推全周富氏算法 ,实现微机距离保护在 段范围内采用“反时限动作特性”;采用网络处理方式消除衰减周期分量、非整次谐波分量等的影响 .仿真结果表明 ,本算法能很有效地减少计算时间 ,并且不受衰减非周期分量、非整次谐波分量等的影响 .另外 ,对所实现的神经网络式微机距离保护也作了简要介绍 ,并给出测试结果 .
This paper chiefly introduces a kind of new method about increasing speed and precision of algorithm for the basic waveforms of signals estimation in distance protection.The method adopts one period window recursive Fourier algorithm to quicken the speed of caculation and reduce the time of calculation;adpots a kind of network handling to remove the influence of attenuation nonperiod parts and fraction resonance waveforms of singnals.Simulation results show that the algorithm can reduce the burden of calculating efficiently,it does not suffer the influence of attenuation nonperiod parts and fraction resonance waveforms of signals parts.Morever,the paper makes a brief introduction to computerized distance protection of adopting the Artifcial Neural Network.
出处
《湖南大学学报(自然科学版)》
EI
CAS
CSCD
2000年第4期66-71,共6页
Journal of Hunan University:Natural Sciences
关键词
距离保护
误差补偿
神经网络
电力系统
distance protection
recursive fourier algohim
error compensation
artificial neural networks