Real-Time Generation of Time-Optimal Quadrotor Trajectories with Semi-Supervised Seq2Seq Learning

Gilhyun Ryou, Ezra Tal, Sertac Karaman
Proceedings of The 6th Conference on Robot Learning, PMLR 205:1860-1870, 2023.

Abstract

Generating time-optimal quadrotor trajectories is challenging due to the complex dynamics of high-speed, agile flight. In this paper, we propose a data-driven method for real-time time-optimal trajectory generation that is suitable for complicated system models. We utilize a temporal deep neural network with sequence-to-sequence learning to find the optimal trajectories for sequences of a variable number of waypoints. The model is efficiently trained in a semi-supervised manner by combining supervised pretraining using a minimum-snap baseline method with Bayesian optimization and reinforcement learning. Compared to the baseline method, the trained model generates up to 20 % faster trajectories at an order of magnitude less computational cost. The optimized trajectories are evaluated in simulation and real-world flight experiments, where the improvement is further demonstrated.

Cite this Paper

BibTeX
@InProceedings{pmlr-v205-ryou23a, title = {Real-Time Generation of Time-Optimal Quadrotor Trajectories with Semi-Supervised Seq2Seq Learning}, author = {Ryou, Gilhyun and Tal, Ezra and Karaman, Sertac}, booktitle = {Proceedings of The 6th Conference on Robot Learning}, pages = {1860--1870}, year = {2023}, editor = {Liu, Karen and Kulic, Dana and Ichnowski, Jeff}, volume = {205}, series = {Proceedings of Machine Learning Research}, month = {14--18 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v205/ryou23a/ryou23a.pdf}, url = {https://proceedings.mlr.press/v205/ryou23a.html}, abstract = {Generating time-optimal quadrotor trajectories is challenging due to the complex dynamics of high-speed, agile flight. In this paper, we propose a data-driven method for real-time time-optimal trajectory generation that is suitable for complicated system models. We utilize a temporal deep neural network with sequence-to-sequence learning to find the optimal trajectories for sequences of a variable number of waypoints. The model is efficiently trained in a semi-supervised manner by combining supervised pretraining using a minimum-snap baseline method with Bayesian optimization and reinforcement learning. Compared to the baseline method, the trained model generates up to 20 % faster trajectories at an order of magnitude less computational cost. The optimized trajectories are evaluated in simulation and real-world flight experiments, where the improvement is further demonstrated. } }
Endnote
%0 Conference Paper %T Real-Time Generation of Time-Optimal Quadrotor Trajectories with Semi-Supervised Seq2Seq Learning %A Gilhyun Ryou %A Ezra Tal %A Sertac Karaman %B Proceedings of The 6th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2023 %E Karen Liu %E Dana Kulic %E Jeff Ichnowski %F pmlr-v205-ryou23a %I PMLR %P 1860--1870 %U https://proceedings.mlr.press/v205/ryou23a.html %V 205 %X Generating time-optimal quadrotor trajectories is challenging due to the complex dynamics of high-speed, agile flight. In this paper, we propose a data-driven method for real-time time-optimal trajectory generation that is suitable for complicated system models. We utilize a temporal deep neural network with sequence-to-sequence learning to find the optimal trajectories for sequences of a variable number of waypoints. The model is efficiently trained in a semi-supervised manner by combining supervised pretraining using a minimum-snap baseline method with Bayesian optimization and reinforcement learning. Compared to the baseline method, the trained model generates up to 20 % faster trajectories at an order of magnitude less computational cost. The optimized trajectories are evaluated in simulation and real-world flight experiments, where the improvement is further demonstrated.
APA
Ryou, G., Tal, E. & Karaman, S.. (2023). Real-Time Generation of Time-Optimal Quadrotor Trajectories with Semi-Supervised Seq2Seq Learning. Proceedings of The 6th Conference on Robot Learning, in Proceedings of Machine Learning Research 205:1860-1870 Available from https://proceedings.mlr.press/v205/ryou23a.html.

Related Material