Presentation Type
Poster
Faculty Mentor’s Full Name
Jaylene Naylor
Faculty Mentor’s Department
Physics and Astronomy
Abstract / Artist's Statement
Building on current methods for accurately determining wind velocity using quadrotor aircraft dynamics, this research helps lay the groundwork for more cost-effective access to wind data as compared to using traditional technologies such as balloon-borne radiosondes. By constructing a dynamic rigid-body model to describe the aircraft’s response to wind perturbations, we are able to determine wind speed and direction. Data collected from University of Montana’s Weather Quadcopter flights is processed using system identification algorithms resulting in excited motion models which are combined to form the rigid-body model. A differential equation describing the state of the aircraft can then be crafted. Solving this differential equation allows for convergence of the system to an estimated state containing the desired wind values. Model tuning is then performed to reduce the uncertainty in wind speed and direction. Preliminary results indicate agreement of 90% between the model and reference data. The ease with which the rigid-body model can be identified for different quadrotor configurations makes it a highly accessible and scalable method for wind determination.
Category
Physical Sciences
Poster Slides
Wind Determination Using a Rigid-Body Model of Quadrotor Motion
Building on current methods for accurately determining wind velocity using quadrotor aircraft dynamics, this research helps lay the groundwork for more cost-effective access to wind data as compared to using traditional technologies such as balloon-borne radiosondes. By constructing a dynamic rigid-body model to describe the aircraft’s response to wind perturbations, we are able to determine wind speed and direction. Data collected from University of Montana’s Weather Quadcopter flights is processed using system identification algorithms resulting in excited motion models which are combined to form the rigid-body model. A differential equation describing the state of the aircraft can then be crafted. Solving this differential equation allows for convergence of the system to an estimated state containing the desired wind values. Model tuning is then performed to reduce the uncertainty in wind speed and direction. Preliminary results indicate agreement of 90% between the model and reference data. The ease with which the rigid-body model can be identified for different quadrotor configurations makes it a highly accessible and scalable method for wind determination.