Long term global path planning for stratospheric airships under time-sequential uncertainty wind fields

Stratospheric airships serve as crucial platforms for near-space applications, and their path planning in complex wind fields is a key challenge limiting future large-scale applications. The wind field exhibits temporal variability and uncertainty, leading to discrepancies between the currently available forecast data and the actual wind field. In this study, we construct determined, undetermined and Time-Sequnential uncertainty wind fields by sequentially updating the global forecast wind field at appropriate frequencies and incorporating uncertainty distributions. Building upon local static field path planning using Markov Decision Processes (MDP), this paper introduces a Sequential Multi-stage Markov Decision Process (DR-MDP) algorithm to find the shortest-time path from the current position to a target area in complex, time-varying wind fields, enabling global long-term path planning.Through simulation experiments, this study analyzes the regional reachability and optimal path selection of aerostats with different propulsion capabilities under three wind field models. Results demonstrate that the proposed method can plan the shortest-time path to a target point in a given two-dimensional wind field environment and provide expected arrival times at each position, providing a feasibility basis for the actual flight planning of the stratospheric airship.