Bayesian Estimation of Power Law Function in Non-homogeneous Poisson Process Applied in Mosul Gas Power Plant – Iraq

Abstract

Non-homogeneous Poisson process with power law intensity function has often been used as a model for describing the failure pattern of repairable systems. Maximum likelihood and Bayesian estimation are used to estimate model parameters. Simulation and realistic application are used and represented by shutting down the gas power plant in Mosul. Stops in hours are designed with the power law random process model in order to obtain a model that represents the average stop time of the units throughout the study period in the best way. The results of the application on the data of the three concerned stations show that the Bayes estimate is better than the maximum likelihood estimate. This proves that the Bayes methods are very accurate and effective in estimating the rate of occurrence parameters.