Sistema de detecção de falta em máquina de indução trífasica aplicando sistemas inteligentes

Abstract

The purpose of this paper is to develop an intelligent system for the monitoring of induction motors capable of detection low intensity short circuits in the machine’s stator. Reducing the number of corrective maintenances in favor of predictive maintenances, therefore making the overall cost of the production plant cheaper, by reducing the occurrence of unscheduled stops in the production line. For that goal a database was created using the results of a computer simulation that emulates the behavior of induction machines under various conditions of fault, with that data various layouts of neural networks wore trained, and their performance was analyzed and compared between each other. This work also discusses about how induction motors operate and the failures that can occur during their operation, the effects of short circuits in the motor’s stator, the 4.0 industry and its implementation in Brazil, the characteristics and applications of intelligent systems and how neural networks work and how they can be trained.