Meteorological path planning using dynamic programming for a solar-powered UAV
Lukas Wirth, Philipp Oettershagen, Jacques Ambühl, Roland Siegwart · Repository for Publications and Research Data (ETH Zurich) · 2015
Solar-powered Unmanned Aerial Vehicles (SUAV) designed for Low-Altitude Long-Endurance (LALE) applications provide potential multi-day continuous flight capability, but are generally prone to local meteorological impediments such as rain, strong winds or reduced solar irradiance.This paper therefore presents METPASS, the Meteorology-aware Trajectory Planning and Analysis Software for Solar-powered UAVs.METPASS optimizes largescale solar-powered UAV missions using a detailed consideration of meteorological effects: An optimal trajectory is found on a 3-D grid for given departure and arrival points by applying a Dynamic Programming approach and a cost function that considers environmental hazards, winds, solar radiation, aircraft parameters and flight time.The cost function is evaluated based on a kinematic and energetic UAV system model and forecast data from the European Centre for Medium-Range Weather Forecasts (ECMWF).The trajectoryplanning environment is applied to an envisioned fully autonomous and solar-powered crossing of the North Atlantic Ocean by AtlantikSolar, a 5.6m-wingspan SUAV developed at ETH Zurich.Results based on historical ECMWF weather data from 2012 and 2013 show that properly pre-optimized routes allow the Atlantic crossing even in case of significant global cloud coverage and that optimal routes can reduce the required flight time by up to 50% (from 106h to 52h) by exploiting wind conditions.