Application of wavelet multi-resolution analysis & perceptron neural networks for classification of transients on transmission line

摘要: This paper proposes a technique that uses Wavelet Multiresolution Analysis (MRA) and Neural Networks for the detection classification of transients in power system. Daubechies eight (db 8) wavelet transforms phase current on transmission line fed from both ends are used. The 5th level output MRA detail signal is used to train perceptron neural network. After training, network able classify all three types system correctly. All work carried out MATLAB Power System Block set program. simulation results show proposed method simple, accurate reliable automate procedure transients. focused identification but can also be easily extended other solutions such as fault location so forth.