Tamper detection using neuro-fuzzy logic [static energy meters]

This paper presents the application of a neuro-fuzzy model to tamper detection in static energy meters. TDNF software has been developed in “C” for the online detection of tamper events in static energy meters. This software provides an option to learn from the input pattern of electrical parameters during various field conditions. A fuzzy membership function is used to account for flexible boundary conditions of electrical parameters. Based upon the training of the neural network, it provides a log of temper events in a real-time domain. The results obtained through simulation are compared with practical conditions. It is observed that information provided by the proposed model is more meaningful as compared to that obtained through existing tamper detection algorithms