An information system for monitoring of power quality disturbances

The paper presents a neural-fuzzy technique-based clarifier for pattern recognition problems with uncertain distributions. Neural networks in the architecture of Frequency Sensitive Competitive Learning and Learning Vector Quantization are first employed to evaluate the decision boundaries separating different patterns to be classified. To deal with the uncertainties of the involved recognition problems, however, the output of the neural networks is used to activate a fuzzy-associative-memory rule-base to accomplish the classification, instead of being taken directly as the final identification. With the Internet and the developed classifiers, an information system can be built up for power quality monitoring over whole power networks.