Image-based adaptive neural control of underactuated aerial mobile robot without direct position measurement

This paper proposes a new image-based adaptive neural controller for the position tracking control of the underactuated aerial mobile robots without the realtime position measurement. To realize this point, a relationship between the position tracking error and the image projection error is first established. One challenging difficulty in stabilization is the fact that the dynamics of the aerial robot is physically underactuated. To address this challenge, a new designed methodology is proposed in this work. In addition, the proposed adaptive controller does not require the explicit inertia information and has a simplified structure because of the inclusion of a new inertia estimator and an optimized structure neural network. Based on the Lyapunov synthesis, the asymptotic convergence of the position tracking error as well as the image projection error to an adjustable region of zero is proved. Lastly, the performance results are provided to verify the effectiveness of the proposed adaptive control scenario.