Study of path planning method for under-actuated blimp-type UAV in stochastic wind disturbance via augmented-MDP

Abstract

Blimp-type unmanned aerial vehicles (BUAVs) do not consume any energy in keeping their longitudinal position in the air, so they are anticipated as a platform that works near the ground but does not touch the ground (for example, for mine search missions). In the path planning of a BUAV flying near the ground, the effects of wind disturbance on the BUAV's motion must be well considered. We have studied path planning of BUAVs via an application of the Markov decision process, and have solved some of unique problems posed by the special characteristics of BUAVs. However, the effects of stochastic wind disturbance have not been sufficiently considered yet in our previous work. To consider the effects of stochastic wind disturbance in the path planning of BUAVs, we studied an application of the augmented Markov decision process (A-MDP), which is a subset of the partially observable MDP. The main issues in the application of A-MDP to path planning of BUAV are defining an efficient state space to reduce calculation cost and consideration of the difference of the wind disturbance effect on BUAV motion caused by its heading angle. A comparison of the results of path generation by proposed method and the previous MDP-based method shows the advantage of the proposed A-MDP-based one.