Improving quadrotor trajectory tracking by replacing PD critics with fuzzy critics in an adaptive critic-based neurofuzzy controller
نویسندگان
1 Mechanical Engineering Department, Khatam Al-Anbia Air Defense University, Tehran, Iran
2 Mechanical Engineering Department, Khatam Al-Anbia Air Defense University, Tehran, Iran
3 Mechanical Engineering Department, Amirkabir University of Technology, Tehran, Iran
doi
10.22111/ijfs.2025.51627.9127چکیده
Quadrotors are increasingly employed in a wide range of applications, including surveying, agriculture, military operations, and rescue missions, due to their superior maneuverability. However, achieving accurate and stable control continues to pose a significant challenge. This study aims to enhance the performance of an adaptive critic-based neurofuzzy controller for quadrotors by replacing the conventional proportional-derivative (PD) critics with fuzzy critics. The control performance was quantitatively evaluated using the root mean square error (RMSE) criterion. Simulation results demonstrate that incorporating fuzzy critics improves performance by more than 39% compared to the traditional PD-based approach. To evaluate the trajectory tracking performance of the adaptive fuzzy-critic-based neurofuzzy controller under disturbance, external disturbances were introduced into the quadrotor model. The controller employing fuzzy critics exhibited strong disturbance rejection, maintaining accurate trajectory tracking despite external interferences. These findings highlight the effectiveness of fuzzy-critic-based neurofuzzy controllers in improving quadrotor trajectory tracking performance under dynamic and uncertain real-world conditions.
