Torque Ripple Reduction of Induction Motor-based Electric Vehicle Using Artificial Neural Network

Globally, the trend is toward reducing CO
2
emissions, and one way to do so is to switch from internal combustion to electric vehicles. To this aim, the electric drive system, which is the most critical component of an electric vehicle, has gained prominence and developed into a significant research issue. The induction motor is an excellent contender for electric vehicle applications because of its simplicity, robustness, and low maintenance. However, the conventional direct torque control scheme has a torque ripple issue, due to the use of hysteresis comparators and the availability of restricted voltage vectors. In this paper, the hysteresis comparator along with the switching table is replaced by an artificial neural network-based controller. The results indicate a significant reduction in torque ripple and improvement in dynamic performance.