A fuzzy Petri-net expert system for power quality classification using wavelet transform

This paper proposes a fuzzy Petri-net (FPNs) expert system for classification of the power quality (PQ) problems. Based on the wavelet transform, the proposed expert system detects and localizes in time the transients of the signals due to various fault events. By calculating related indices to the PQ, the FPNs inference-module draws possible causes of the disturbances as well as its credibility values from its knowledge base. Attributed to the FPNs structure, the knowledge expression and inference process can be achieved systematically and graphically. To verify the performance of the proposed FPNs expert system, a wide variety of disturbance signals simulated by Electromagnetic Transients Program (EMTP) package were employed. The development and the effectiveness of the proposed system as well as the testing results are presented in this paper.