Speed Control of BLDC Motor Using Neural Network Based Model Reference Adaptive Control (MRAC) For Electric Vehicle Drive (In the Case of Yaris Car)

Abstract

Electric vehicles are widely considered a viable solution to reduce fossil oil dependence and environmental footprints in the ground transportation sector. Electric vehicle reduce oil consumption and air pollutant emission where concerns about oil security and availability and the negative environmental impact of petroleum-based transportation systems. Different Motors are being used in Electric Vehicles. Brushless DC motors showed best performance as compared to other motors like(brushed DC motor, Induction motor, Brushless AC motor and other motor), because of its great efficiency, high energy density, Less maintenance, Long operating life, Low electric noise, Better speed versus torque characteristics and etc. Different researchers have proposed different controller model in order to control speed of Brushless DC Motors. Among these controllers, Model reference adaptive controller is the one which is a component controller that emulates the reference model behavior efficiently. Conventional MRAC scheme is used for linear systems. However, since BLDC is nonlinear system by nature using conventional MRAC is very difficult. Hence in this thesis it proposed Neural Network based MRAC to control the speed of BLDC motors to compensate its nonlinearity which is not considered in conventional MRAC. The proposed neural network based model reference adaptive controller can significantly improve the system behavior and force the system to follow the reference model and minimize the error between the model and the plant output. Adaptive law using Lyapunov stability criteria for updating the controller parameters online have been formulated. The simulation results show that the proposed neural network-based MRAC has 0.0236 sec rise time, 0.0450 sec settling time and overshoot 0.0736 percent .This performance is closely the same with model reference performance which has 0.0257sec rise time, 0.0438sec settling time and overshoot 0.0363 percent The performance of the neural network based MRAC have been compared with conventional MRAC and MRAC-PID. The neural network based MRAC was selected for speed control of BLDC motor drive system. Because of in the case of NN based MRAC the output speed is closely the same with model reference output and the difference between them is very small when it compared with other controller which is listed in this thesis. The result have been using MATLAB/ SIMULINK