Speed control of induction motors using dynamical neural network

ABSTRACTMaster ThesisSPEED CONTROL OF INDUCTION MOTORS USINGDYNAMICAL NEURAL NETWORKKoray ŞAHİNFırat UniversityGraduate School of Natural and Applied SciencesDepartment of Electronics and Computer Education2006, Page: 70Induction motors are widely used in industry since they have lower maintenancerequirements, lower cost and the ability to operate in dirty and explosive environment. Highperformance electrical drives used for variable speed applications should have; an acceptablespeed and position control performance in transient and steady state, a good response for thevariable speed demands and robustness against the load and parameter variations of the motor.Due to the nonlinear and complex structure of induction motors, simple and effective speedcontrol system is required for the induction motor drives.In this thesis, speed control of an induction motor is implemented using the dynamicalneural network (DNN). NN inputs are selected as the speed error and the change of the errorand, the NN output is the torque current. NN is trained using the speed error. Locally recurrentNN structure is used as a speed controller and control performance is compared with thefeedforward counterpart. Speed control of induction motor including the trained NN controlleris simulated with MATLAB-Simulink. Simulation results showing the performance of thecontrol system are presented under the load and parameter variations.Keywords: Induction motors, dynamical neural networks, speed control.