The recognition of power quality events by analyzing voltage and current waveform disturbances is a very important task for power system monitoring. This paper presents a new approach to the recognition of power quality disturbances using wavelet transforms and neural networks. The proposed method employs wavelet transform multiresolution signal decomposition techniques together with multiple neural networks using a learning vector quantization network as a powerful classifier. Various transient events are tested, such as voltage sag, voltage swell, interruption, notching, impulsive transient, and harmonic distortion. The results show that the classifier can efficiently detect and classify different types of power quality disturbance.
展开▼
