A neural network approach for acoustic leak monitoring in the VVER440 pressure vessel head

A neural network approach has been developed for localizing leakages and estimating the leak rate in the VVER-440 pressure vessel head. Results are presented from experiments with simulated leaks. Three-layer perceptron networks were found to be best suited for leak localization and for the estimation of leak rates. However, the estimation of leak rates required an additional neural network because a different normalization procedure was necessary for extracting features from RMS values of the acoustic emission sensors. Perceptron networks with continuously valued outputs corresponding to the coordinates of the leak positions were useful for indentifying even leak positions which had not been offered during training.