Estimación de parámetros de un motor de inducción usando redes neuronales y sistemas neurodifusos

This thesis presents the estimation of parameters of an induction motor using neural networks and neurofuzzy systems; such parameters are the rotor ux, the position of the rotor ux and the motor speed. These parameters are indispensable for a vector control scheme. The methodology is based on the se...

Descripción completa

Detalles Bibliográficos
Autor: Miguel Antonio Cuautle Pat
Tipo de recurso: tesis de maestría
Estado:Versión aceptada para publicación
Fecha de publicación:2018
País:México
Institución:Instituto Nacional de Astrofísica, Óptica y Electrónica
Repositorio:Repositorio Institucional del INAOE
Idioma:español
OAI Identifier:oai:inaoe.repositorioinstitucional.mx:1009/2029
Acceso en línea:http://inaoe.repositorioinstitucional.mx/jspui/handle/1009/2029
Access Level:acceso abierto
Palabra clave:info:eu-repo/classification/Inspec/Estimate
info:eu-repo/classification/Inspec/Induction motor
info:eu-repo/classification/Inspec/Parameters
info:eu-repo/classification/Inspec/Neural network
info:eu-repo/classification/Inspec/Vivado HLS
info:eu-repo/classification/Inspec/ANFIS
info:eu-repo/classification/cti/1
info:eu-repo/classification/cti/22
info:eu-repo/classification/cti/2203
Descripción
Sumario:This thesis presents the estimation of parameters of an induction motor using neural networks and neurofuzzy systems; such parameters are the rotor ux, the position of the rotor ux and the motor speed. These parameters are indispensable for a vector control scheme. The methodology is based on the search of architectures that can estimate these parameters through the training of neural networks and neurofuzzy systems. The selection of the architectures is based on a heuristic reasoning, making tests and observing the behavior of the root mean square error. Once the architectures were obtained, they were described in High-Level Synthesis, using Vivado HLS, a tool of Xilinx. This description was accomplished in C++, where results of the use of hardware utilization and latency are presented, for a later implementation on a FPGA. The validation of the architectures was analyzed by System Generator, which allows the modeling in Simulink in order to explore the operation without the use of tracing the design to hardware.