Artificial Neural Network Control of Vector Controlled Induction Motor

Many researches have been carried out to induction motors for starting, braking, speed reversal and speed control, because they are relatively cheap, reliable and rugged machines due to absence of slip rings or commutators. Induction motors exhibit highly coupled, nonlinear time varying system which is difficult to control since some state variables are difficult to be measured. However, recent advancements in semiconductor power electronics and microcontrollers have made it possible to use induction motors in many variable speed drive applications, as they are capable of similar performance as DC motor. This paper presents a study for indirect vector control of induction motor as it can be operated over a wide speed range, including low speed, with rapid, accurate torque control and good momentary overload capabilities. Also, the use of artificial neural network, ANN is proposed to emulate the function of Indirect-Field-Oriented-Control (IFOC), to perform the critical function of synchronous speed estimation internally, transformation from three-phase ABC currents to two-phase d-q synchronous frame currents and transformation from twophase d-q synchronous frame voltages to three-phase ABC voltages during both constant torque and constant power regions, also for motor reversing and braking modes.