[4] WU H, SONG S J, YOU K Y et al. Depth control of model-free AUVs via reinforcement learning[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 49, 2499-2510(2019).

[9] PALOMERAS N, EL-FAKDI A, CARRERAS M et al. COLA2: a control architecture for AUVs[J]. IEEE Journal of Oceanic Engineering, 37, 695-716(2012).

[10] [10] BALL P J, LU C, PARKERHOLDER J, et al. Augmented wld models facilitate zeroshot dynamics generalization from a single offline environment[C]Proceedings of the 38th International Conference on Machine Learning. PMLR, 2021: 619629.

[12] [12] PINTO L, DAVIDSON J, SUKTHANKAR R, et al. Robust adversarial reinfcement learning[C]Proceedings of the 34th International Conference on Machine Learning. Sydney, Australia: PMLR, 2017.

[13] [13] DENNIS M, JAQUES N, VINITSKY E, et al. Emergent complexity zeroshot transfer via unsupervised environment design[C]Proceedings of the 34th International Conference on Neural Infmation Processing Systems. Vancouver: Curran Associates Inc. , 2020.

[14] [14] KAKADE S, LANGFD J. Approximately optimal approximate reinfcement learning[C]Proceedings of the Nieenth International Conference on Machine Learning. San Francisco: Mgan Kaufmann Publishers Inc. , 2002.

[15] [15] SCHULMAN J, WOLSKI F, DHARIWAL P, et al. Proximal policy optimization algithms[JOL]. arXiv: 1707.06347, 2017 [20240121]. http:arxiv.gabs1707.06347.

[16] [16] FOSSEN T I. Marine control systems–guidance. navigation, control of ships, rigs underwater vehicles[R]. Trondheim: Marine Cyberics, 2002.

[17] [17] YU W H, TAN J, KAREN LIU C, et al. Preparing f the unknown: learning a universal policy with online system identification[JOL]. Robotics: Science Systems, 2017 [20240123]. https:tml.stanfd.edupublications2017preparingunknownlearninguniversalpolicyonlinesystemidentification.

[18] [18] ZHOU W X, PINTO L, GUPTA A. Environment probing interaction policies[C]7th International Conference on Learning Representations. New leans: ICLR, 2019.

[19] [19] LEE K, SEO Y, LEE S, et al. Contextaware dynamics model f generalization in modelbased reinfcement learning[C]Proceedings of the 37th International Conference on Machine Learning. 2020: 5757−5766.

[20] VON BENZON M, SØRENSEN F F, UTH E et al. An open-source benchmark simulator: control of a BlueROV2 underwater robot[J]. Journal of Marine Science and Engineering, 10, 1898(2022).