The Simulation of the Reliabity of Assets Using the Monte Carlo Method. The Specific-Case of Complex and Coherent Systems K-Out-of-N, With Censored Data

Management of the life cycle of physical assets is based on the estimated life of the equipment, whether during the project, installation and start-up or throughout its useful life. In many situations, engineering needs to calculate or simulate equipment's estimated useful life using accurate data, which is often censored or incomplete. In many cases, equipment, regarding failures, can be structurally represented as block diagram systems. For this article, complex and coherent systems (from a reliability point of view) were studied. Censored data usually results in the loss of important information but must be included in the reliability analysis models of these complex systems. The article develops the complex and coherent systems theory and the respective reliability models. With a new and original approach, simulation algorithms were developed to generate random and censored data (right-censored and type I data). Two case studies of complex systems were designed to validate the algorithms. For each case, a set of simulations was developed with the variation of the reliability models' different parameters to compare better, tune and optimize the simulation of these complex systems. One of the relevant results shows that the more censored data in the sample, the greater the bias and error about the true value. The variation of the parameter β (shape factor) from β = 0:5 to β = 1:5 proportionally increases the bias. This article aims to validate the use of the Monte Carlo simulation tool and the Weibull statistical distribution and contribute to improving, with more precision and speed, algorithms for simulating the reliability of complex and coherent systems in the presence of censored data.