Metaheuristic-optimized PID, LQR, and H∞ controllers for quadrotor UAV attitude control

Quadrotor UAVs are nonlinear, underactuated, and strongly coupled, rendering attitude stabilization particularly challenging in the presence of disturbances. This work compares PID, LQR, and H∞ controllers based on a standard nonlinear quadrotor dynamic model, and uses GA and PSO for gain autotuning. The purpose here is to establish if metaheuristic optimization can outperform conventional MATLAB tuning methods in terms of accuracy, robustness, and efficiency of control. The optimized controllers will be subjected to nominal flight conditions, transient wind-gust disturbances, and actuator saturation. The results demonstrate that PSO yields the lowest tracking error, fastest convergence, and most efficient control effort, while PSO H∞ results in the strongest disturbance rejection capability, and PSO LQR yields smooth, well-damped responses. In conclusion, the results prove that metaheuristic optimization forms a reliable, effective approach for improvement of quadrotor attitude control in the presence of realistic operational constraints..