dc.contributor.author | Havenstrøm, Simen Theie | |
dc.contributor.author | Rasheed, Adil | |
dc.contributor.author | San, Omer | |
dc.date.accessioned | 2021-09-07T05:20:16Z | |
dc.date.available | 2021-09-07T05:20:16Z | |
dc.date.created | 2020-06-18T03:19:42Z | |
dc.date.issued | 2020 | |
dc.identifier.issn | 2296-9144 | |
dc.identifier.uri | https://hdl.handle.net/11250/2773870 | |
dc.description.abstract | In this article, we explore the feasibility of applying proximal policy optimization, a state-of-the-art deep reinforcement learning algorithm for continuous control tasks, on the dual-objective problem of controlling an underactuated autonomous surface vehicle to follow an a priori known path while avoiding collisions with non-moving obstacles along the way. The AI agent, which is equipped with multiple rangefinder sensors for obstacle detection, is trained and evaluated in a challenging, stochastically generated simulation environment based on the OpenAI gym Python toolkit. Notably, the agent is provided with real-time insight into its own reward function, allowing it to dynamically adapt its guidance strategy. Depending on its strategy, which ranges from radical path-adherence to radical obstacle avoidance, the trained agent achieves an episodic success rate close to 100%. | en_US |
dc.language.iso | eng | en_US |
dc.publisher | Institute of Electrical and Electronics Engineers (IEEE) | en_US |
dc.relation.uri | https://arxiv.org/abs/2006.09792 | |
dc.rights | Navngivelse 4.0 Internasjonal | * |
dc.rights.uri | http://creativecommons.org/licenses/by/4.0/deed.no | * |
dc.title | Deep Reinforcement Learning Controller for 3D Path-following and Collision Avoidance by Autonomous Underwater Vehicles | en_US |
dc.type | Peer reviewed | en_US |
dc.type | Journal article | en_US |
dc.description.version | publishedVersion | en_US |
dc.source.journal | Frontiers in Robotics and AI | en_US |
dc.identifier.doi | 10.1109/access.2020.2976586 | |
dc.identifier.cristin | 1816044 | |
cristin.ispublished | true | |
cristin.fulltext | preprint | |
cristin.qualitycode | 1 | |