Abstract
This paper investigates a nonparametric modeling scheme for underwater vehicles to achieve continuous-time dynamics modeling, which is essential for various marine missions, control, and navigation of these vehicles. The proposed scheme addresses the challenges posed by the nonlinearity, strong coupling, and complex structure of underwater vehicles through local adaptive moment estimation Gaussian processes learning. This approach constructs mappings between hydrodynamics and motion states while providing uncertainty estimates of the dynamics model. A local weighted strategy is used to construct local models to localize Gaussian processes learning, and an adaptive moment estimation method is designed using gradients of innovation to tune hyperparameters of Gaussian processes automatically. Moreover, a subspace index is created and updated based on feature distance measures to improve the computational efficiency of Gaussian processes learning in each local model. The developed scheme can perform real-time simulation considering environmental disturbances and is applied to a 6 degree-of-freedom autonomous underwater vehicle. The results demonstrate that this scheme is an effective mathematical modeling tool for underwater vehicles dynamics.
Similar content being viewed by others
Data availability
Enquiries about data availability should be directed to the authors.
References
Wu, H., Niu, W., Wang, S., Yan, S.: An optimization method for control parameters of underwater gliders considering energy consumption and motion accuracy. Appl. Math. Model. 90, 1099–1119 (2021). https://doi.org/10.1016/j.apm.2020.10.015
Liu, X., Zhang, M., Yao, F., Chu, Z.: Observer-based region tracking control for underwater vehicles without velocity measurement. Nonlinear Dyn. 108(4), 3543–3560 (2022). https://doi.org/10.1007/s11071-022-07393-8
Shen, C., Shi, Y.: Distributed implementation of nonlinear model predictive control for AUV trajectory tracking. Automatica 115, 108863 (2020). https://doi.org/10.1016/j.automatica.2020.108863
Liang, X., Li, Y., Peng, Z., Zhang, J.: Nonlinear dynamics modeling and performance prediction for underactuated AUV with fins. Nonlinear Dyn. 84(1, SI), 237–249 (2016). https://doi.org/10.1007/s11071-015-2442-1
Gibson, S.B., Stilwell, D.J.: Hydrodynamic parameter estimation for autonomous underwater vehicles. IEEE J. Ocean. Eng. 45(2), 385–394 (2020). https://doi.org/10.1109/JOE.2018.2877489
Ahmed, F., Xiang, X., Jiang, C., Xiang, G., Yang, S.: Survey on traditional and AI based estimation techniques for hydrodynamic coefficients of autonomous underwater vehicle. Ocean Eng. 268, 113300 (2023). https://doi.org/10.1016/j.oceaneng.2022.113300
Nouri, N.M., Valadi, M., Asgharian, J.: Optimal input design for hydrodynamic derivatives estimation of nonlinear dynamic model of AUV. Nonlinear Dyn. 92(2), 139–151 (2018). https://doi.org/10.1007/s11071-017-3611-1
Randeni, S., Forrest, A.L., Cossu, R., Leong, Z.Q., Ranmuthugala, D., Schmidt, V.: Parameter identification of a nonlinear model: replicating the motion response of an autonomous underwater vehicle for dynamic environments. Nonlinear Dyn. 91(2), 1229–1247 (2018). https://doi.org/10.1007/s11071-017-3941-z
Gartner, N., Richier, M., Dune, C., Hugel, V.: Hydrodynamic parameters estimation using varying forces and numerical integration fitting method. IEEE Robot. Autom. Lett. 7(4), 11713–11719 (2022). https://doi.org/10.1109/LRA.2022.3205126
Cardenas, P., Barros, E.A.: Estimation of AUV hydrodynamic coefficients using analytical and system identification approaches. IEEE J. Ocean. Eng. 45(4), 1157–1176 (2020). https://doi.org/10.1109/JOE.2019.2930421
Deng, F., Levi, C., Yin, H., Duan, M.: Identification of an autonomous underwater vehicle hydrodynamic model using three Kalman filters. Ocean Eng. 229, 108962 (2021). https://doi.org/10.1016/j.oceaneng.2021.108962
Sabet, M.T., Daniali, H.M., Fathi, A., Alizadeh, E.: Identification of an autonomous underwater vehicle hydrodynamic model using the extended, cubature, and transformed unscented Kalman filter. IEEE J. Ocean. Eng. 43(2), 457–467 (2018). https://doi.org/10.1109/JOE.2017.2694470
Xu, H., Hinostroza, M.A., Wang, Z., Guedes Soares, C.: Experimental investigation of shallow water effect on vessel steering model using system identification method. Ocean Eng. 199, 106940 (2020). https://doi.org/10.1016/j.oceaneng.2020.106940
Song, S., Liu, J., Guo, J., Zhang, C., Yang, T., Cui, J.: Efficient velocity estimation and location prediction in underwater acoustic sensor networks. IEEE Internet Things J. 9(4), 2984–2998 (2022). https://doi.org/10.1109/JIOT.2021.3094305
Jing, G., Lei, L., Gang, Y.: Dynamic modeling and experimental analysis of an underwater glider in the ocean. Appl. Math. Model. 108, 392–407 (2022). https://doi.org/10.1016/j.apm.2022.03.034
Duan, K., Fong, S., Chen, C.L.P.: Multilayer neural networks-based control of underwater vehicles with uncertain dynamics and disturbances. Nonlinear Dyn. 100(4), 3555–3573 (2020). https://doi.org/10.1007/s11071-020-05720-5
Bande, M., Wehbe, B.: online model adaptation of autonomous underwater vehicles with LSTM Networks. In: OCEANS 2021: San Diego-Porto, pp. 1–6 (2021). https://doi.org/10.23919/OCEANS44145.2021.9705897
Wang, D., Wan, J., Shen, Y., Qin, P., He, B.: Hyperparameter optimization for the LSTM method of AUV model identification based on Q-learning. J. Mar. Sci. Eng. 10, 1002 (2022). https://doi.org/10.3390/jmse10081002
Xu, F., Zou, Z.-J., Yin, J.-C., Cao, J.: Identification modeling of underwater vehicles’ nonlinear dynamics based on support vector machines. Ocean Eng. 67, 68–76 (2013). https://doi.org/10.1016/j.oceaneng.2013.02.006
Pei, T., Yu, C., Zhong, Y., Lian, L.: Adaptive event-triggered mechanism-based online system identification framework for marine craft. Ocean Eng. 278, 114572 (2023). https://doi.org/10.1016/j.oceaneng.2023.114572
Ramirez, W.A., Kocijan, J., Leong, Z.Q., Nguyen, H.D., Jayasinghe, S.G.: Dynamic system identification of underwater vehicles using multi-output Gaussian processes. Int. J. Autom. Comput. 18, 681–693 (2021). https://doi.org/10.1007/s11633-021-1308-x
Schuerch, M., Azzimonti, D., Benavoli, A., Zaffalon, M.: Correlated product of experts for sparse Gaussian process regression. Mach. Learn. 112(5), 1411–1432 (2023). https://doi.org/10.1007/s10994-022-06297-3
Yang, K., Lu, J., Wan, W., Zhang, G., Hou, L.: Transfer learning based on sparse gaussian process for regression. Inf. Sci. 605, 286–300 (2022). https://doi.org/10.1016/j.ins.2022.05.028
Bender, M., Tian, L., Fan, X., Kurdila, A., Mueller, R.: Spatially recursive estimation and gaussian process dynamic models of bat flapping flight. Nonlinear Dyn. 95(1), 217–237 (2019). https://doi.org/10.1007/s11071-018-4560-z
Renson, L., Sieber, J., Barton, D.A.W., Shaw, A.D., Neild, S.A.: Numerical continuation in nonlinear experiments using local Gaussian process regression. Nonlinear Dyn. 98(4, SI), 2811–2826 (2019). https://doi.org/10.1007/s11071-019-05118-y
Zhang, Z., Ren, J.: Locally weighted non-parametric modeling of ship maneuvering motion based on sparse Gaussian process. J. Mar. Sci. Eng. (2021). https://doi.org/10.3390/jmse9060606
Yoo, C., Heon Lee, J.J., Anstee, S., Fitch, R.: Path planning in uncertain ocean currents using ensemble forecasts. In: 2021 IEEE International Conference on Robotics and Automation (ICRA), pp. 8323–8329 (2021). https://doi.org/10.1109/ICRA48506.2021.9561626
Fossen, T.I.: Handbook of Marine Craft Hydrodynamics and Motion Control. Wiley, England (2011). https://doi.org/10.1002/9781119994138
McMahan, H.B., Streeter, M.J.: Adaptive bound optimization for online convex optimization. In: Annual Conference Computational Learning Theory (2010). https://doi.org/10.48550/arXiv.1002.4908
Reddi, S.J., Kale, S., Kumar, S.: On the convergence of Adam and beyond. ArXiv (2019) https://doi.org/10.48550/arXiv.1904.09237
Moulay, E., Léchappé, V., Plestan, F.: Properties of the sign gradient descent algorithms. Inf. Sci. 492, 29–39 (2019). https://doi.org/10.1016/j.ins.2019.04.012
Safaryan, M.H., Richtárik, P.: Stochastic sign descent methods: new algorithms and better theory. In: ICML (2021). https://doi.org/10.48550/arXiv.1905.12938
Tong, Q., Liang, G., Bi, J.: Calibrating the adaptive learning rate to improve convergence of ADAM. Neurocomputing 481, 333–356 (2022). https://doi.org/10.1016/j.neucom.2022.01.014
Martin, S.C., Whitcomb, L.L.: Experimental identification of six-degree-of-freedom coupled dynamic plant models for underwater robot vehicles. IEEE J. Ocean. Eng. 39(4), 662–671 (2014). https://doi.org/10.1109/JOE.2013.2280492
Healey, A.J., Good, M.R.: The NPS AUV II autonomous underwater vehicle testbed: Design and experimental verification. Nav. Eng. J. 104(3), 191–202 (1992). https://doi.org/10.1111/j.1559-3584.1992.tb02238.x
Zhang, Z., Ren, J., Bai, W.: MIMO non-parametric modeling of ship maneuvering motion for marine simulator using adaptive moment estimation locally weighted learning. Ocean Eng. 261, 112103 (2022). https://doi.org/10.1016/j.oceaneng.2022.112103
Acknowledgements
The authors would like to appreciate the anonymous reviewers for their constructive suggestions which comprehensively improve the quality of the paper. This work was partially supported by the National Natural Science Foundation of China (Grant Nos. 51779029, 61903092, 51939001) and Natural Science Foundation of Sichuan Province (Grant No. 2022NSFSC0891).
Author information
Authors and Affiliations
Contributions
ZZ involved in conceptualization, methodology, software, simulation, visualization, writing original draft preparation. JR took part in conceptualization, writing review and editing, supervision, project administration, funding acquisition.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A
Proof of Theorem 1
In order to make the following statement clearer, as mentioned in Theorem 1, here we will take one-dimensional space as an example to illustrate, i.e., \(\varvec{\vartheta }_{k,i} \rightarrow \vartheta _k, \varvec{m}_{k,i} \rightarrow m_k, \varvec{n}_{k,i} \rightarrow n_k\), etc. Based on the above formulation of ‘adaptive learning problem’ and Eq. (33), \({\vartheta }_{k+1}\) is constructed as follows
From Lemma 2, using \(u_1 = {\vartheta }_{k+1}\), \(u_2 = {\vartheta }_{*}\), \(x={\vartheta }\) and \(\mathcal {E} = n_k^{1/2}\), we have:
Since \({\mathcal {A}}_k\) is the k-th iteration element about \({\mathcal {A}}\), the term \([(\beta _{1} m_{k-1}+(1-\beta _{1}){\mathcal {A}}_k)({\vartheta }_k-{\vartheta }_*)]\) approximates to \([k{{g}}_k({\vartheta }_k-{\vartheta }_*)]\). Note that \(\beta _{1} \in [0,1)\) and \(\beta _{2} \in [0,1)\), and using Cauchy–Schwartz and Young’s inequality rearranging inequality Eq. (A2), we have:
where Cauchy–Schwartz and Young’s inequality is \(ab \le a^2\epsilon /2 + b^2 / 2\epsilon \), \(\forall \epsilon >0\).
According to the purpose in ‘Adaptive learning problem,’ by convexity of \(f(\bullet )\) and Eq. (A3), we have:
Since \(\beta _{1} \in [0,1)\) and \(\beta _{2} \in [0,1)\), \(0 \le n_{k-1} \le n_{k}\),
where for \({\sum }_{t=1}^T \mathcal {Q}(n_k^{-1 / 4} m_k)^2\) in Eq. (A5), assuming \(0< c < n_k\), \(\forall k \in [\mathcal {K}]\), according to the definition of \(m_k\) in Eq. (33), then
According to Cauchy–Schwartz inequality \(\langle x, y \rangle \le \Vert x\Vert \Vert y\Vert \), and \({\sum }_{k=1}^\mathcal {K}1/\sqrt{k} \le 1+\log \mathcal {K}\), then Eq. (A6) becomes
Apply Eqs. (A7) to (A5), we have:
It can be seen from Eq. (A8) that \(f_k({\vartheta }_k)-f_k({\vartheta }_*)\) is converged, which means \(\lim _{k \rightarrow \infty } \left[ f_k({\vartheta }_k)-f_k({\vartheta }_*)\right] ^2=0\) can be obtained. For the assumed convex case, Theorem 1 means that the convergence time complexity of AME is \(O(\sqrt{\mathcal {K}})\). \(\square \)
Appendix B
To supplement the details of the process in Fig. 2, the pseudocode of our proposed LAGPL algorithm is presented here as shown in Algorithm 1, which includes the training and prediction stages.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhang, Z., Ren, J. Nonparametric dynamics modeling for underwater vehicles using local adaptive moment estimation Gaussian processes learning. Nonlinear Dyn 112, 5477–5502 (2024). https://doi.org/10.1007/s11071-024-09314-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-024-09314-3