Title: Enhancing VTOL control using twin delayed deep deterministic policy gradient-based controller

Authors: Haitham M. Al-Radhi; Khaled A. El-Metwally

Addresses: Department of Electrical Power Engineering, Faculty of Engineering, Cairo University, Giza, Egypt ' Department of Electrical Power Engineering, Faculty of Engineering, Cairo University, Giza, Egypt

Abstract: This study presents the application of a twin delayed deep deterministic policy gradient (TD3) based controller for vertical take-off and landing (VTOL) system control. The TD3 algorithm is a deep reinforcement learning (DRL) algorithm (actor-critic type) designed for continuous control problems. The DRL agent learns to optimise actions to maximise rewards by interacting with the environment. VTOL aircraft is a nonlinear system that shows high complexity with variable aerodynamic parameters during the flight. The proposed controller is implemented on a one-degree-of-freedom VTOL system. The performance of the proposed controller is compared to that of a PID controller and a deep deterministic policy gradient (DDPG) based controller, another DRL algorithm. The evaluation includes analysing step response characteristics and the sum square of errors. MATLAB and Simulink are utilised for the implementation and analysis. The results indicate that the TD3-based controller exhibits performance better than the PID controller with reduced settling time and free overshoot.

Keywords: vertical take-off and landing; VTOL control; reinforcement learning; AI; deep learning; TD3-based controller; PID controller; actor-critic method; MATLAB; deep reinforcement learning; DRL; deep deterministic policy gradient; DDPG.

DOI: 10.1504/IJISTA.2024.143255

International Journal of Intelligent Systems Technologies and Applications, 2024 Vol.22 No.4, pp.378 - 404

Received: 02 Jul 2023
Accepted: 02 Nov 2023
Published online: 11 Dec 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article