A Hybrid Hidden Markov Model for Pipeline Leakage Detection

In this paper, a deep neural network hidden Markov model (DNN-HMM) is proposed to detect pipeline leakage location. A long divided into several sections and the occurs in different section that defined as state of (HMM). The hybrid HMM, i.e., DNN-HMM, consists (DNN) with multiple layers exploit non-linear data. DNN initialized by using belief (DBN). DBN pre-trained built stacking top-down restricted Boltzmann machines (RBM) compute emission probabilities for HMM instead Gaussian mixture (GMM). Two comparative studies based on numbers states model-hidden (GMM-HMM) DNN-HMM are performed. accuracy testing performance between detected sequence actual measured micro F1 score. score approaches 0.94 GMM-HMM method it close 0.95 when three sections. experiment divides five sections, 0.69, while 0.96 method. results demonstrate can learn better data achieve compared