Leak Detection of Pipeline Using a Hybrid of Neural-Adaptive Tabu Search Algorithm

This paper presents a new hybrid of Neural-Adaptive Tabu Search (NATS) for leakage detection in pipelines. The proposed cooperative algorithms are formed from Artificial Neural Network (ANN) and Adaptive Tabu Search (ATS). The article shows comparison studies of the ANN and NATS algorithms. The search performance evaluation is performed on the standard benchmark from University of California at Irvine (UCI) Machine Learning Repository. The experiment uses water leakage signals from a field-test yard. The results show that the leaking pipeline can be efficiently detected.