Modeling and control of a brushless direct current motor

In this thesis, the performance analysis of the obtained system was performed by applying proportional-integral (PI), proportional-integral-derivative (PID) and fuzzy logic-PI (FGPI) controllers to a brushless direct current (DC) motor model. For this purpose, firstly, the system model of this brushless direct current motor was created in MATLAB Simulink environment with the appropriate parameter values selected. Then, the parameters of PI, PID and fuzzy logic-PI (FGPI) controllers were optimized with nine different metaheuristic optimization algorithms such as classical Ziegler-Nichols method and dragonfly optimization algorithm (DA), water cycle algorithm (WCA), particle swarm optimization (PSO), genetic algorithm (GA), gray wolf optimization (GWO), artificial bee colony (ABC), atom search optimization (ASO), bat optimization algorithm (BA) and flower pollination optimization algorithm (FPA) and applied to our brushless DC motor model and compared according to the rise times, settling times and percent overshoot values of the system output signals. In addition, the convergence curves and operating speeds of the objective functions of the nine metaheuristic optimization methods used are compared with each other. When the performance criteria of our brushless DC motor system with optimized PI and FGPI controllers are compared in positive and negative regions under unit step input, it is seen that FGPI controllers reduce percentage overshoots and increase rise times compared to PI controllers. When PID and FGPI controllers optimized with GWO and PSO are applied to our brushless DC motor model, it is seen that the calculated settling times are smaller than those of other metaheuristic algorithms and Ziegler-Nichols method. Since speed control is important for brushless DC motors, GWO and PSO algorithms that give the smallest settling times gain importance in this regard. When the overshoot percentages for PID, PI and FGPI optimized with metaheuristic algorithms are taken into consideration, it is seen that using PID controller in our motor model provides less percentage overshoots and generally smaller rise times compared to the other two controllers. When the convergence curves of metaheuristic algorithms are compared for both PID and PI controllers, it is understood that the GWO algorithm generally gives lower error values. When the running times of metaheuristic algorithms are compared for both PID and PI controllers, it is seen that the FPA algorithm reaches the optimization results faster.