Goudinakis, George

A new methodology was developed for flow regime identification in pipes. The method utilizes the pattern recognition abilities of Artificial Neural Networks and the unprocessed time series of a system-monitoring-signal. The methodology was tested with synthetic data from a conce…