Neural-Swarm: Decentralized Close-Proximity Multirotor Control Using Learned Interactions
Abstract
In this paper, we present Neural-Swarm, a nonlinear decentralized stable controller for close-proximity flight of multirotor swarms. Close-proximity control is challenging due to the complex aerodynamic interaction effects between multirotors, such as downwash from higher vehicles to lower ones. Conventional methods often fail to properly capture these interaction effects, resulting in controllers that must maintain large safety distances between vehicles, and thus are not capable of close-proximity flight. Our approach combines a nominal dynamics model with a regularized permutation-invariant Deep Neural Network (DNN) that accurately learns the high-order multi-vehicle interactions. We design a stable nonlinear tracking controller using the learned model. Experimental results demonstrate that the proposed controller significantly outperforms a baseline nonlinear tracking controller with up to four times smaller worst-case height tracking errors. We also empirically demonstrate the ability of our learned model to generalize to larger swarm sizes.
Additional Information
© 2020 IEEE. We thank Michael O'Connell and Xichen Shi for helpful discussions and Anya Vinogradsky for help with the firmware implementation. The work is funded in part by Caltech's Center for Autonomous Systems and Technologies (CAST) and the Raytheon Company.Attached Files
Submitted - 2003.02992.pdf
Submitted - ICRA_2020_Neural_Swarm.pdf
Files
Name | Size | Download all |
---|---|---|
md5:039dc7e516635cfeed8700f3d726d57d
|
2.7 MB | Preview Download |
md5:3685c5608e2f7cd11bea8fee2dc5fc15
|
2.7 MB | Preview Download |
Additional details
- Eprint ID
- 99548
- Resolver ID
- CaltechAUTHORS:20191029-155055963
- Center for Autonomous Systems and Technologies (CAST)
- Raytheon Company
- Created
-
2019-10-29Created from EPrint's datestamp field
- Updated
-
2021-11-16Created from EPrint's last_modified field
- Caltech groups
- GALCIT, Center for Autonomous Systems and Technologies (CAST)