Design of Deep Learning Controller for Vector Controlled Induction Motor Drive

This paper tends to the utilization of Deep Learning concept to structure the controller, to investigate achievability of using deep learning into control issues. Induction motors play an important role in industrial applications. If speed is not controlled properly, it is for all intents and purposes difficult to accomplish the preferred task for precise application. Induction motors are known for their simplicity and reliability. They are maintenance free and low-cost electric drives. Due to the lack of ability of traditional control techniques such as PI and PID controllers to serve under wide scope of activity, AI controllers are broadly utilized in industries such as fuzzy logic controller, neuro-fuzzy controller (NFC), genetic algorithm. By learning PI controller, which is mostly utilized in the industries, the suggested Deep Learning-based controller is planned. The input and output of PI controller are utilized as learning informational index for Deep Learning (DL) system. To stay away from computational weight, just speed error is given as the input to the proposed deep learning controller, not at all like a conventional controller which utilizes speed error and its derivative. Deep belief network (DBN) control process is utilized to design the DL controller. DLC gives incredible speed exactness. By using MATLAB Simulink, simulation is carried out. Results of suggested DLC and NFC are compared; finally, it was led to show the effectiveness and usefulness of the suggested procedure.