Horizontal trajectory control of stratospheric airships in wind field using Q-learning algorithm

Abstract

This paper proposes an adaptive horizontal trajectory control method for stratospheric airships in uncertain wind field using Q-learning algorithm. Firstly, horizontal trajectory control of the airships is decomposed into the target tracking, and the observation model of airships is constructed. Then, the Markov decision process (MDP) model of airships is established, in which the action strategy is determined by the wind direction, and a cerebellar model articulation controller (CMAC) neural network is designed to optimize the action strategy for each state. Finally, numerical simulations demonstrate that the proposed control method performs well stability and intelligent decision-making ability in the process of horizontal trajectory control for stratospheric airships.