Perturbation-mitigated USV Navigation with Distributionally Robust Reinforcement Learning
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917113341411328 |
|---|---|
| author | Zhang, Zhaofan Yang, Minghao Xie, Sihong Xiong, Hui |
| author_facet | Zhang, Zhaofan Yang, Minghao Xie, Sihong Xiong, Hui |
| contents | The robustness of Unmanned Surface Vehicles (USV) is crucial when facing unknown and complex marine environments, especially when heteroscedastic observational noise poses significant challenges to sensor-based navigation tasks. Recently, Distributional Reinforcement Learning (DistRL) has shown promising results in some challenging autonomous navigation tasks without prior environmental information. However, these methods overlook situations where noise patterns vary across different environmental conditions, hindering safe navigation and disrupting the learning of value functions. To address the problem, we propose DRIQN to integrate Distributionally Robust Optimization (DRO) with implicit quantile networks to optimize worst-case performance under natural environmental conditions. Leveraging explicit subgroup modeling in the replay buffer, DRIQN incorporates heterogeneous noise sources and target robustness-critical scenarios. Experimental results based on the risk-sensitive environment demonstrate that DRIQN significantly outperforms state-of-the-art methods, achieving +13.51\% success rate, -12.28\% collision rate and +35.46\% for time saving, +27.99\% for energy saving, compared with the runner-up. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_00030 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Perturbation-mitigated USV Navigation with Distributionally Robust Reinforcement Learning Zhang, Zhaofan Yang, Minghao Xie, Sihong Xiong, Hui Robotics Artificial Intelligence The robustness of Unmanned Surface Vehicles (USV) is crucial when facing unknown and complex marine environments, especially when heteroscedastic observational noise poses significant challenges to sensor-based navigation tasks. Recently, Distributional Reinforcement Learning (DistRL) has shown promising results in some challenging autonomous navigation tasks without prior environmental information. However, these methods overlook situations where noise patterns vary across different environmental conditions, hindering safe navigation and disrupting the learning of value functions. To address the problem, we propose DRIQN to integrate Distributionally Robust Optimization (DRO) with implicit quantile networks to optimize worst-case performance under natural environmental conditions. Leveraging explicit subgroup modeling in the replay buffer, DRIQN incorporates heterogeneous noise sources and target robustness-critical scenarios. Experimental results based on the risk-sensitive environment demonstrate that DRIQN significantly outperforms state-of-the-art methods, achieving +13.51\% success rate, -12.28\% collision rate and +35.46\% for time saving, +27.99\% for energy saving, compared with the runner-up. |
| title | Perturbation-mitigated USV Navigation with Distributionally Robust Reinforcement Learning |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2512.00030 |