ROS package for nonlinear model predictive controller (NMPC). The resulting NMPC tracks a reference position trajectory. Besides, it expects the learned values for the disturbance parameters
wind_nmhe_regression: A nonlinear moving horizon estimation-based regression package.
wind_gp_regression: A Gaussian process-based regression package.\
Moreover, the following packages can be used to command the reference trajectory:
m100_trajectory: A ROS package to generate reference trajectory to traverse a wall.
trajectories: A ROS package to generate reference trajectories.
Note: the utilization of the disturbance parameters can be switched off by setting "use_dist_estimates" to "false" within the launch file.
This NMPC package is utilized in the following work. Please don't forget to consider citing it if you use these codes in your work.
Plain text:
M. Mehndiratta and E. Kayacan, "Gaussian Process-based Learning Control of Aerial Robots for Precise Visualization of Geological Outcrops," 2020 European Control Conference (ECC), Saint Petersburg, Russia, 2020, pp. 10-16.
Bibtex:
@INPROCEEDINGS{9143655,
author={M. {Mehndiratta} and E. {Kayacan}},
booktitle={2020 European Control Conference (ECC)},
title={Gaussian Process-based Learning Control of Aerial Robots for Precise Visualization of Geological Outcrops},
year={2020},
volume={},
number={},
pages={10-16}
}