New solutions are required as forest fires overwhelm our current fire management capacity all over the globe. Wind speed and wind direction heavily determine both the spread and intensity of these wildfires while the information available for firefighters is very limited. Meteorology data is often communicated once per hour and does not factor in the complex topology or other local aerodynamic disturbances.
This proposed project aimed to investigate the use of drones to solve the problem of lacking meteorology information available while in any wind-sensitive situations. Two different approaches to estimate the speed and direction were compared, one existing, and one original. Both use data available from common on-board sensors to estimate wind with a calibrated estimator in simulation with real recorded wind data.
Our original approach to estimating wind speed and direction produced more accurate results for all wind samples tested. It did, however, need a much larger area to operate in, requiring the immediate environment around it to first be cleared from possible obstacles.