Um aplicativo para a execução de sistemas especialistas no planejamento e controle da manutenção

• Ano de defesa: 2012
• Autor(a) principal: Malinowski, Erich Lacerda
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Ponta Grossa; Programa de Pós-Graduação em Engenharia de Produção
• Palavras-chave em Português: Sistemas especialistas (Computação); Fábricas - Manutenção; Lógica difusa; Inteligência artificial; Expert systems (Computer science); Plant maintenance; Fuzzy logic; Artificial intelligence
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/1462

In this study, a development of an application for the execution of Expert Systems based on the Fuzzy Logic is carried out in order to help managers on the diagnosis of the operation of equipments and on the determination of the best moment for doing intervention or maintenance. The research had as motivation the necessity to overcome some characteristics of the Expert System DEMOI, developed and executed in the tool Fuzzy Logic Toolbox of the environment Matlab™. So, a tool in language JAVA and C was developed aiming to get an improvement in the processing time of the informations and the integration between Expert Systems of monitoring and planning of maintenance operations. For developing this research were considered preliminary tests of operation of the Expert Systems linked to Matlab software, as well as the development steps of the proposed application. The results of the simulations, carried out on a practical case of analysis of vibration of a rotary machine, showed the benefits of the new application, principally when the processing time was evaluated. In this sense, a considerable reduction in time of return for diagnoses with elevated quantities of variables and inferences rules was possible by using the new tool. Additionally, with the introduction of the application was provided an environment of integration between Expert Systems of monitoring and planning of the maintenance, which can be used for managers without the limitations present in applications of commercial softwares.