Classification of Power Transient Signals Using Probabilistic Neural Network

This paper uses probabilistic neural networks and wavelet transforms to give an in-depth analysis of transient categorization in power systems. This technique, which derives from the Bayes rule, attempts to estimate the underlying probability density functions that are necessary for the Bayes rule but are frequently not easily accessible. Based on multi-resolution analysis, feature vectors for different power quality disturbances are extracted using the wavelet transform. This method's mathematical underpinnings are presented in the paper, and then applied for classification of power quality disturbances. Wavelet analysis of transient waveforms are employed in this study to derive features. Results of PNN model are compared with standard real time simulated data which are generated using PSCAD/EMTDC software.