本文开发了一种基于人工神经元网络(ANN)和快速付里叶变换(FFT)技术进行预想事故快速筛选的方法。利用快速解耦潮流计算的迭代一次法(1P－1Q)分别构造了反映预想事故严重程度的有功性能指标PIp和无功性能指标PIv，同时还构造了一个多层感知型神经元网络并用BP算法加以训练.神经元网络的输入经过快速付氏变换可以大大提高网络的训练速度.算例表明，本算法具有较高的性能指标计算精度，且性能指标的构造避免了遮蔽现象的发生，同时ANN的特点也使得预想事故的筛选速度大为提高。%A new approach based on artificial neural network (ANN) combinedwith fast Fourier transform (FFT) techniques is developed for single-line contingency screening in steady- state security analysis. The results obtained from 1P-1Q iteration of the fast decoupled load flow calculation are adopted to construct two kinds of performance indices PIp (active performance index) and PIv (reactive performance index) which reflect the severe degrees of certain contingencies. A multi-layered ANN is trained to calculate the performance indices using error back propagation algorithm. FFT for inputs of the ANN is used to improve and speed up the training procedure. The effectiveness of the proposed method is demonstrated by contingency screening on IEEE test systems. The calculating accuracy ,high capturing rate and analysis speed for contingency screening are obtained using the proposed method.
展开▼
