Modelling and Optimization of the Preventive Maintenance of an Air Compressor using the Intensity Reduction Model

摘要

Modeling the reliability of complex repairable systems (CRS) is often intended to be used as a decision support tool for planning the best periodicity of preventive maintenance (PM) of systems. However, modeling with the renewal process (RP) and the non-homogeneous Poisson process (NHPP) considering respectively the case of perfect (PR) and minimal (MR) repair can put the intervention on the system either too soon or too late by deviating from its best relevant periodicity. In this article, we consider the intensity reduction model (IRM) for an approach to optimize the frequency of preventive maintenance (PM) of an air compressor that has operated for nearly three years in the LPG (liquefied petroleum gas) company of NAFTAL (subsidiary of the SONATRACH Group). The estimation of the parameters of this model, making it possible to highlight the effect of preventive maintenance actions (PM), is carried out by the maximum likelihood approach (MLE) which allows to develop, on the programming language MATLAB, an algorithm for simulating the future performance of air compressors. The objective is to propose, over a fixed time horizon, the best frequency of preventive maintenance based on the economic criterion representing the average cost generated by each recommended PM interval. This realistic approach will ensure a balance between preventive and corrective maintenance actions.