Data-driven modeling for flapping flight
- Motivation: difficult to generate flapping flight models because of complexities of flapping-wing aerodynamics
- Created data set by recording load cell data and free-flight data of the bat robot B2
- Created a parameter estimation method using a combination of first-principles and data-driven tools to generate a model that improves long-term flight prediction
- Demonstrated significantly improved long-term flight prediction with the new model
- Presented this work at the RSS 2020 virtual robotics conference