Direct Torque Controller of SRM for EV Application Based on Neural Network

Direct Torque Controller of SRM for EV Application Based on Neural Network

연구 분야: Infrastructure

학회: International Conference on Broadband Communications, Networks and Systems

This paper presents the machine learning based control of cascaded converter fed SRM drive of electric vehicle. The electric vehicle which is driven by a 6/4 Switched Reluctance Motor (SRM) powered by four battery banks is considered in this paper. The new topology of converter is proposed to drive SRM effectively for application of electric car. The direct torque controller is implemented with the help of neural network for effective speed controller with minimum ripples in electromagnetic torque. The required pulses are generated with space vector Pulse Width Modulation (PWM) technique. The electromagnetic torque generated by SRM needs to be maintained at ripples free for smooth operation of electric car. The mathematical validation is implemented to achieve the required power rating of SRM for Toyota Car. The proposed topology of converter has a facility of using four battery banks, hence the charging time of batteries will be minimized. The direct torque controller with model reference adaptive controller is implemented for establishing sensor less operation. The proposed system is implemented in MATLAB/Simulink. The extensive results are presented which validate the proposed system for steady state and transient state requirements of the drive.