A Stochastic Model for Predicting the Number of Failures in Gas Systems

Abstract

We propose a stochastic model based on a nonhomogeneous Poisson process (NHPP) for predicting the number of failures. The new NHPP-based stochastic model uses an intensity function consisting of trend and periodic parts. This allows us to precisely predict failures in gas systems in Japan because of the flexibility of the NHPP and specific characters of our proposed intensity function. We show the proposed model well predicts the number of failures in a term in comparison with two prominent stochastic models and in terms of prediction error.