FPGA-embedded Fuzzy Logic-based Processor for online Detection of Broken Bars on Induction Motors

Cesar da Costa¹; Mauro Hugo Mathias²; Christian Oliveira santin³

¹ IFSP- Instituto Federal de Educação, Ciência e Tecnologia; ² UNESP-Universidade Estadual Paulista Julio de Mesquita; ³ IFSP- Instituto Federal de Educação, Ciência e Tecnologia, 01109-010, São Paulo, São Paulo, Brasil

doi:10.20906/CPS/COB-2015-0008

Resumo

Monitoring, fault detection and diagnosis of electric induction motors are becoming more and more important issues in the field of electrical machines as new data processing techniques and new methods for analyzing the stator current of an induction motor. Machine monitoring plays a crucial role in detecting incipient faults and, in this way, preventing maintenance process is now a critical success factor in the industrial environment, and the ideal combination of techniques and procedures, as well as the implementation of new analysis tools and diagnostic equipment with new technologies, must be reviewed constantly. Special attention has been devoted to non-invasive methods, which are capable to detect faults using measured data without disassembling the machine and its structural parts. The fuzzy logic approach may help diagnose induction motor faults. The fuzzy system can identify motor condition with high accuracy. In this study, we present an example of fusion between the embedded software and hardware in an FPGA-based processor. The embedded architecture consists of a real-time processor, field programmable gate array (FPGA) chipset, and I/O modules. Fuzzy system has two inputs and three outputs. The numerical amplitude values of the two fault frequency components of the spectrum stator current are intended as input to the fuzzy system and are converted into linguistic information by using fuzzy subsets and their corresponding membership functions. Experimental results are presented in terms of motor fault detection accuracy and knowledge extraction feasibility. Preliminary results show that the fuzzy logic-based processor described herein can be used for accurately detecting broken bars in induction motors. Keywords: Rotating electrical machine, condition monitoring, diagnostic, digital signal processing

Palavras-chave: Rotating electrical machine; condition monitoring; digital signal processing