Physics-Inspired Temporal Learning of Quadrotor Dynamics for Accurate Model Predictive Trajectory Tracking
Please be aware that this code was originally implemented for research purposes and may be subject to changes and any fitness for a particular purpose is disclaimed. To inquire about commercial licenses, please contact Alessandro Saviolo (alessandro.saviolo@nyu.edu), Guanrui Li (lguanrui@nyu.edu), and Prof. Giuseppe Loianno (loiannog@nyu.edu).
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <https://www.gnu.org/licenses/>.
If you publish a paper with our data, please cite our paper:
@article{saviolo2022pitcn,
author = {Saviolo, Alessandro and Li, Guanrui and Loianno, Giuseppe},
title = {Physics-Inspired Temporal Learning of Quadrotor Dynamics for Accurate Model Predictive Trajectory Tracking},
journal = {IEEE Robotics and Automation Letters},
year = {2022},
volume = {7},
number = {4},
pages = {10256-10263},
doi = {10.1109/LRA.2022.3192609}
}
Accurately modeling quadrotor's system dynamics is critical for guaranteeing agile, safe, and stable navigation. The model needs to capture the system behavior in multiple flight regimes and operating conditions, including those producing highly nonlinear effects such as aerodynamic forces and torques, rotor interactions, or possible system configuration modifications. Classical approaches rely on handcrafted models and struggle to generalize and scale to capture these effects. In this paper, we present a novel Physics-Inspired Temporal Convolutional Network (PI-TCN) approach to learning quadrotor's system dynamics purely from robot experience. Our approach combines the expressive power of sparse temporal convolutions and dense feed-forward connections to make accurate system predictions. In addition, physics constraints are embedded in the training process to facilitate the network's generalization capabilities to data outside the training distribution. Finally, we design a model predictive control approach that incorporates the learned dynamics for accurate closed-loop trajectory tracking fully exploiting the learned model predictions in a receding horizon fashion. Experimental results demonstrate that our approach accurately extracts the structure of the quadrotor's dynamics from data, capturing effects that would remain hidden to classical approaches. To the best of our knowledge, this is the first time physics-inspired deep learning is successfully applied to temporal convolutional networks and to the system identification task, while concurrently enabling predictive control.
We release our collected dataset which we used for training and evaluating the Physics-Inspired Temporal Convolutional Network.
The data was collected by controlling the quadrotor in a series of flights in an indoor environment 10x6x4 meters at the Agile Robotics and Perception Lab (ARPL) at the New York University. The environment is equipped with a Vicon motion capture system that allows recording accurate position and attitude measurements at 100Hz. Additionally, we record the onboard motor speeds.
The dataset consists of 68 trajectories with a total of 58'03'' flight time. The trajectories range from straight-line accelerations to circular motions, but also parabolic maneuvers and lemniscate trajectories. All the trajectories are performed for any axis combination (i.e., x-y, x-z, y-z) and with different speeds and accelerations.
We provide the dataset as a set of bag files, each representing one trajectory, and pdf illustrations of the raw collected data.
The bag files can be downloaded here, while the pdf files can downloaded here.
We provide a python script for processing the bag files. The script loads each bag file, splines and syncronizes selected topics (odometry from Vicon, angular velocity from IMU and rotor speeds from ESC), and saves the dataset as .csv files (one for training, multiple for testing). The python script offers the possibility to filter the parsed data to minimize the effect of sensor noise. We provide the code for running low-pass Butterworth and Savitzky–Golay filters.
Process the data by running the python script:
python process_data.py
Note: the downloaded bag files' folder should be stored in the same folder as the python script.