AI-Based Control of a Novel Airfoil-Shaped Quadrotor Hybrid Airship

This research aims to control an airfoil-shaped quadrotor hybrid airship through the utilization of mathematical modeling and machine learning techniques. The primary objective of this study is to enhance the aircraft’s maneuverability, stability, and energy efficiency while addressing the inherent challenges associated with its control. To achieve precise maneuvering and adherence to target parameters, a comprehensive mathematical model was integrated with a neural network-based control system. The selected configuration incorporates tilting rotors as the primary source of thrust, accompanied by horizontal stabilizers for pitch control. The research investigates the optimal control of the rotors and stabilizers to reach stability and the desired behavior of the airship. Machine learning techniques, specifically neural networks, are employed to achieve adaptable and robust control. The methodology involves constructing a mathematical model of the airship, training the neural network using data generated by the model, and exploring various neural network architectures and control methods. The outcomes of this research contribute to the advancement of airship technology, facilitating the development of cost-effective, environmentally friendly autonomous airships with a wide range of applications. Future work will focus on expanding the mathematical model, establishing a simulation environment, and conducting field tests to validate the performance of the developed control system.