Modelos com reparo imperfeito no R

• Ano de defesa: 2015
• Autor(a) principal: Rafael da Silva Fernandes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: il; Engenharia de produção
• Link de acesso: http://hdl.handle.net/1843/BUBD-A45M3Y

Most organizations face the problem of having high costs associated with the maintenance activity. When equipment failures occur leading to unscheduled stoppages for corrective maintenance, the associated costs include those related to the unavailability of equipment, namely, cost of spare parts, labor costs and the consequences of unscheduled outages of the production line. On the other hand, even when the company executes a plan of preventive maintenance, in which the stoppages for repair are scheduled, there are costs incurred by this activity. The goal of any maintenance policy is, therefore, the implementation of practices that lead to a reductions of the expected total cost of this activity. In this sense, an analysis of the historical data of system failures is necessary. It is the necessary to search for probabilistic/statistical models to implement such analysis in order to provide valuable information for decision making such as, the choice of an appropriate maintenance policy. Probabilistic/statistical models for the analysis of repairable systems have been widely discussed in the literature. The most common probabilistic models for repairable systems, the Renewal Process and the Non-homogeneous Poisson Process, deal with the assumptions of perfect repair (which brings the system to a state as good as new - AGAN) and minimum repair (state as bad as old) after failure, respectively. However, such assumptions are not realistic for a number of practical situations since a repair can bring a given system to an intermediate level between the two above mentioned ones. Several authors studied the behavior of systems subjected to this type of repair, known as imperfect repair and among the models proposed are, for example, the class ofARA (Arithmetic Reduction of Age) and ARI (Arithmetic Reduction of Intensity) of Doyen e Gaudoin (2004).. One difficulty for the disseminations of such models, very useful in practice is the absence of a library of computational routines. This work provides students, researchers, engineers or anyone interested in the subject, a detailed user guide to the library (Imperfect Repair), developed in R environment. This tool is practical, free and easy to use for reliability analysis of repairable systems, in particular, under the imperfect repair condition.