| Summary: | Unmanned aerial systems (UAS) are increasingly used for delivery operations due to its high mobility, low cost and ease of deployment. However, most commercial UAS are battery powered with limited energy for flight and there are different dynamic factors that give rise to uncertainty, thus accurate energy consumption estimation is key to manage delivery operation. This project developed a data-driven approach to estimate energy consumption using the dynamic states of UAS, such as velocity, acceleration, wind and payload. The energy model was implemented based on existing models from literature review. The methodology was then extended to generalize the energy model for different multirotor UAS through outlining key steps for data preprocessing, model training, optimization and validation. Possible application of the energy model includes optimization of delivery routes and providing energy consumption information to drone operators and unmanned traffic management.
|
|---|