Modelagem e otimização da manutenção preventiva imperfeita com fator de melhoria variável e severidades independentes

• Ano de defesa: 2024
• Autor(a) principal: Santos, Uerviton Silva dos
• Orientador(a): Pereira, Fabio Henrique
• Banca de defesa: Pereira, Fabio Henrique , Paschoalin Filho, João Alexandre , Librantz, Andre Felipe Henriques , Belan, Peterson Adriano
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação em Informática e Gestão do Conhecimento
• Departamento: Informática
• País: Brasil
• Palavras-chave em Português: política de manutenção; manutenção preventiva imperfeita; custo de manutenção; algoritmo genético; otimização por enxame de partículas
• Palavras-chave em Inglês: maintenance policy; imperfect preventive maintenance; maintenance cost; genetic algorithm; particle swarm optimization
• Área do conhecimento CNPq: CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO
• Link de acesso: http://bibliotecatede.uninove.br/handle/tede/3522

A maintenance policy aims to plan the cycle of maintenance actions necessary to maintain the reliability and full operation of a system. Several studies address the identification of the optimal number of maintenance actions within a planning horizon, but without considering the possibility of variation in the severity rates of these actions. However, this work proposes an approach that identifies the optimal number of preventive maintenance actions within an observed time interval, the timing for each maintenance to be performed, and the respective severities assigned to the maintenance actions. To achieve this objective, the methodology employed suggests a first step with the adjustment of the reliability model based on system failure data, considering the failure intensity as a function of time. And, in a second step, the minimization of a cost function that takes into account the estimated costs of preventive and corrective maintenance for the period. For optimizations, two methods were used: the first one was the Genetic Algorithm (AG) and the second was the Particle Swarm Optimization (PSO). The approach included experiments conducted in four test scenarios, considering variable systems and severities, in which it was possible to evaluate the effectiveness of the function in obtaining the number of maintenance actions for the observed period, the optimal times, and the respective severity rates. A comparison between the AG and PSO algorithms was carried out to identify the best optimization approach, and PSO proved to be superior to AG.