Comparison of ANN- and GA-based DTC eCAR

Abstract

In this paper, an artificial intelligence (AI)-integrated direct torque control (DTC) scheme is developed for an electric vehicle (EV or eCAR) propulsion motor drive. In addition, a comparison is made between adaptive neural network (ANN) and genetic algorithm (GA)-based torque controllers. The integration of AI into EVs has attracted the attention of many researchers in terns if drive control, dynamic stability, speed estimation, and energy management strategies. Amidst the various motor drive control strategies, DTC schemes with space vector pulse width modulation (SVPWM) have gained prominence due to its fast torque (speed) control capability. The smooth control of a DTC-eCAR propulsion motor is accomplished by the use of AI algorithms. The applications of ANN and GA algorithms for tuning the torque controller are tested and the behavior of an eCAR in terms of drive range, percentage of state of charge (SOC), and energy consumption for different driving conditions is observed using MATLAB simulations.