Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
126 tokens/sec
GPT-4o
47 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Port-Hamiltonian Neural ODE Networks on Lie Groups For Robot Dynamics Learning and Control (2401.09520v2)

Published 17 Jan 2024 in cs.RO, cs.SY, and eess.SY

Abstract: Accurate models of robot dynamics are critical for safe and stable control and generalization to novel operational conditions. Hand-designed models, however, may be insufficiently accurate, even after careful parameter tuning. This motivates the use of machine learning techniques to approximate the robot dynamics over a training set of state-control trajectories. The dynamics of many robots are described in terms of their generalized coordinates on a matrix Lie group, e.g. on $SE(3)$ for ground, aerial, and underwater vehicles, and generalized velocity, and satisfy conservation of energy principles. This paper proposes a port-Hamiltonian formulation over a Lie group of the structure of a neural ordinary differential equation (ODE) network to approximate the robot dynamics. In contrast to a black-box ODE network, our formulation embeds energy conservation principle and Lie group's constraints in the dynamics model and explicitly accounts for energy-dissipation effect such as friction and drag forces in the dynamics model. We develop energy shaping and damping injection control for the learned, potentially under-actuated Hamiltonian dynamics to enable a unified approach for stabilization and trajectory tracking with various robot platforms.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (63)
  1. L. Ljung, “System identification,” Wiley encyclopedia of electrical and electronics engineering, 1999.
  2. D. Nguyen-Tuong and J. Peters, “Model learning for robot control: a survey,” Cognitive processing, vol. 12, no. 4, 2011.
  3. M. Deisenroth and C. E. Rasmussen, “PILCO: A model-based and data-efficient approach to policy search,” in International Conference on Machine Learning (ICML), 2011.
  4. G. Williams, N. Wagener, B. Goldfain, P. Drews, J. M. Rehg, B. Boots, and E. A. Theodorou, “Information theoretic MPC for model-based reinforcement learning,” in IEEE International Conference on Robotics and Automation (ICRA), 2017.
  5. M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Multistep neural networks for data-driven discovery of nonlinear dynamical systems,” arXiv preprint arXiv:1801.01236, 2018.
  6. K. Chua, R. Calandra, R. McAllister, and S. Levine, “Deep reinforcement learning in a handful of trials using probabilistic dynamics models,” in Advances in Neural Information Processing Systems (NeurIPS), 2018.
  7. M. Lutter and J. Peters, “Combining physics and deep learning to learn continuous-time dynamics models,” The International Journal of Robotics Research, vol. 42, no. 3, pp. 83–107, 2023. [Online]. Available: https://doi.org/10.1177/02783649231169492
  8. J. K. Gupta, K. Menda, Z. Manchester, and M. J. Kochenderfer, “A general framework for structured learning of mechanical systems,” arXiv preprint arXiv:1902.08705, 2019.
  9. M. Cranmer, S. Greydanus, S. Hoyer, P. Battaglia, D. Spergel, and S. Ho, “Lagrangian neural networks,” in ICLR 2020 Workshop on Integration of Deep Neural Models and Differential Equations, 2020.
  10. S. Greydanus, M. Dzamba, and J. Yosinski, “Hamiltonian neural networks,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 32, 2019.
  11. Z. Chen, J. Zhang, M. Arjovsky, and L. Bottou, “Symplectic recurrent neural networks,” International Conference on Learning Representations (ICLR), 2020.
  12. M. A. Roehrl, T. A. Runkler, V. Brandtstetter, M. Tokic, and S. Obermayer, “Modeling system dynamics with physics-informed neural networks based on Lagrangian mechanics,” arXiv preprint arXiv:2005.14617, 2020.
  13. Y. Lu, S. Lin, G. Chen, and J. Pan, “ModLaNets: Learning generalisable dynamics via modularity and physical inductive bias,” in Proceedings of the 39th International Conference on Machine Learning, ser. Proceedings of Machine Learning Research, K. Chaudhuri, S. Jegelka, L. Song, C. Szepesvari, G. Niu, and S. Sabato, Eds., vol. 162.   PMLR, 2022.
  14. C. Neary and U. Topcu, “Compositional learning of dynamical system models using port-hamiltonian neural networks,” in Learning for Dynamics and Control Conference.   PMLR, 2023, pp. 679–691.
  15. A. Sanchez-Gonzalez, N. Heess, J. T. Springenberg, J. Merel, M. Riedmiller, R. Hadsell, and P. Battaglia, “Graph networks as learnable physics engines for inference and control,” in International Conference on Machine Learning (ICML), 2018.
  16. M. Lutter, C. Ritter, and J. Peters, “Deep Lagrangian networks: using physics as model prior for deep learning,” in International Conference on Learning Representations (ICLR), 2018.
  17. T. Bertalan, F. Dietrich, I. Mezić, and I. G. Kevrekidis, “On learning Hamiltonian systems from data,” Chaos: An Interdisciplinary Journal of Nonlinear Science, vol. 29, no. 12, 2019.
  18. M. Finzi, K. A. Wang, and A. G. Wilson, “Simplifying Hamiltonian and Lagrangian neural networks via explicit constraints,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 33, 2020.
  19. Y. D. Zhong, B. Dey, and A. Chakraborty, “Symplectic ODE-Net: learning Hamiltonian dynamics with control,” in International Conference on Learning Representations (ICLR), 2019.
  20. J. Willard, X. Jia, S. Xu, M. Steinbach, and V. Kumar, “Integrating physics-based modeling with machine learning: A survey,” arXiv preprint arXiv:2003.04919, 2020.
  21. A. J. Maciejewski, “Hamiltonian formalism for euler parameters,” Celestial Mechanics, vol. 37, no. 1, 1985.
  22. R. Shivarama and E. P. Fahrenthold, “Hamilton’s equations with Euler parameters for rigid body dynamics modeling,” Journal of Dynamic Systems, Measurement, and Control, vol. 126, no. 1, 2004.
  23. T. Duong and N. Atanasov, “Hamiltonian-based Neural ODE Networks on the SE(3) Manifold For Dynamics Learning and Control,” in Proceedings of Robotics: Science and Systems, Virtual, July 2021.
  24. R. T. Chen, Y. Rubanova, J. Bettencourt, and D. Duvenaud, “Neural ordinary differential equations,” in Advances in Neural Information Processing Systems (NeurIPS), 2018.
  25. Y. D. Zhong, B. Dey, and A. Chakraborty, “Dissipative SymODEN: Encoding Hamiltonian dynamics with dissipation and control into deep learning,” in ICLR 2020 Workshop on Integration of Deep Neural Models and Differential Equations, 2020.
  26. A. Van Der Schaft and D. Jeltsema, “Port-Hamiltonian systems theory: An introductory overview,” Foundations and Trends in Systems and Control, vol. 1, no. 2-3, 2014.
  27. Y. Song, A. Romero, M. Müller, V. Koltun, and D. Scaramuzza, “Reaching the limit in autonomous racing: Optimal control versus reinforcement learning,” Science Robotics, vol. 8, no. 82, p. eadg1462, 2023. [Online]. Available: https://www.science.org/doi/abs/10.1126/scirobotics.adg1462
  28. T. Salzmann, E. Kaufmann, J. Arrizabalaga, M. Pavone, D. Scaramuzza, and M. Ryll, “Real-time neural mpc: Deep learning model predictive control for quadrotors and agile robotic platforms,” IEEE Robotics and Automation Letters, vol. 8, no. 4, pp. 2397–2404, 2023.
  29. D. Scaramuzza and E. Kaufmann, “Learning agile, vision-based drone flight: From simulation to reality,” in The International Symposium of Robotics Research.   Springer, 2022, pp. 11–18.
  30. A. Loquercio, E. Kaufmann, R. Ranftl, M. Müller, V. Koltun, and D. Scaramuzza, “Learning high-speed flight in the wild,” Science Robotics, vol. 6, no. 59, p. eabg5810, 2021.
  31. J. Ibarz, J. Tan, C. Finn, M. Kalakrishnan, P. Pastor, and S. Levine, “How to train your robot with deep reinforcement learning: lessons we have learned,” The International Journal of Robotics Research, vol. 40, no. 4-5, pp. 698–721, 2021.
  32. T. X. Nghiem, J. Drgoňa, C. Jones, Z. Nagy, R. Schwan, B. Dey, A. Chakrabarty, S. Di Cairano, J. A. Paulson, A. Carron et al., “Physics-informed machine learning for modeling and control of dynamical systems,” arXiv preprint arXiv:2306.13867, 2023.
  33. F. Djeumou, C. Neary, E. Goubault, S. Putot, and U. Topcu, “Neural networks with physics-informed architectures and constraints for dynamical systems modeling,” in Learning for Dynamics and Control Conference.   PMLR, 2022, pp. 263–277.
  34. A. J. Thorpe, C. Neary, F. Djeumou, M. M. Oishi, and U. Topcu, “Physics-informed kernel embeddings: Integrating prior system knowledge with data-driven control,” arXiv preprint arXiv:2301.03565, 2023.
  35. L. Ruthotto and E. Haber, “Deep neural networks motivated by partial differential equations,” Journal of Mathematical Imaging and Vision, 2019.
  36. R. Wang, R. Walters, and R. Yu, “Incorporating symmetry into deep dynamics models for improved generalization,” in International Conference on Learning Representations (ICLR), 2021.
  37. M. Lutter, K. Listmann, and J. Peters, “Deep Lagrangian Networks for end-to-end learning of energy-based control for under-actuated systems,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2019.
  38. T. Beckers, J. Seidman, P. Perdikaris, and G. J. Pappas, “Gaussian process port-hamiltonian systems: Bayesian learning with physics prior,” in 2022 IEEE 61st Conference on Decision and Control (CDC).   IEEE, 2022, pp. 1447–1453.
  39. T. Beckers, “Data-driven Bayesian Control of Port-Hamiltonian Systems,” in IEEE Conference on Decision and Control (CDC), 2023.
  40. P. Toth, D. J. Rezende, A. Jaegle, S. Racanière, A. Botev, and I. Higgins, “Hamiltonian generative networks,” in International Conference on Learning Representations (ICLR), 2019.
  41. J. Mason, C. Allen-Blanchette, N. Zolman, E. Davison, and N. Leonard, “Learning interpretable dynamics from images of a freely rotating 3d rigid body,” arXiv preprint arXiv:2209.11355, 2022.
  42. J. E. Marsden and M. West, “Discrete mechanics and variational integrators,” Acta Numer., vol. 10, pp. 357–514, 2001.
  43. A. Havens and G. Chowdhary, “Forced variational integrator networks for prediction and control of mechanical systems,” in Learning for Dynamics and Control (L4DC), 2021, pp. 1142–1153.
  44. V. Duruisseaux, T. Duong, M. Leok, and N. Atanasov, “Lie Group Forced Variational Integrator Networks for Learning and Control of Robot Systems,” in Learning for Dynamics and Control (L4DC), 2023, pp. 731–744.
  45. R. Chen and M. Tao, “Data-driven prediction of general hamiltonian dynamics via learning exactly-symplectic maps,” in Proceedings of the 38th International Conference on Machine Learning, ser. Proceedings of Machine Learning Research, M. Meila and T. Zhang, Eds., vol. 139.   PMLR, 2021, pp. 1717–1727.
  46. O. So, G. Li, E. A. Theodorou, and M. Tao, “Data-driven discovery of non-newtonian astronomy via learning non-euclidean hamiltonian,” ICML Machine Learning and the Physical Sciences Workshop, 2022.
  47. R. Ortega, M. W. Spong, F. Gómez-Estern, and G. Blankenstein, “Stabilization of a class of underactuated mechanical systems via interconnection and damping assignment,” IEEE Transactions on Automatic Control, vol. 47, no. 8, 2002.
  48. J. Á. Acosta, M. Sanchez, and A. Ollero, “Robust control of underactuated aerial manipulators via IDA-PBC,” in IEEE Conference on Decision and Control (CDC).   IEEE, 2014.
  49. O. B. Cieza and J. Reger, “IDA-PBC for underactuated mechanical systems in implicit Port-Hamiltonian representation,” in European Control Conference (ECC), 2019.
  50. Z. Wang and P. Goldsmith, “Modified energy-balancing-based control for the tracking problem,” IET Control Theory & Applications, vol. 2, no. 4, 2008.
  51. C. Souza, G. V. Raffo, and E. B. Castelan, “Passivity based control of a quadrotor UAV,” IFAC Proceedings Volumes, vol. 47, no. 3, 2014.
  52. L. Furieri, C. L. Galimberti, M. Zakwan, and G. Ferrari-Trecate, “Distributed neural network control with dependability guarantees: a compositional port-hamiltonian approach,” in Learning for Dynamics and Control Conference.   PMLR, 2022, pp. 571–583.
  53. C. L. Galimberti, L. Furieri, L. Xu, and G. Ferrari-Trecate, “Hamiltonian deep neural networks guaranteeing nonvanishing gradients by design,” IEEE Transactions on Automatic Control, vol. 68, no. 5, pp. 3155–3162, 2023.
  54. E. Sebastián, T. Duong, N. Atanasov, E. Montijano, and C. Sagüés, “LEMURS: Learning distributed multi-robot interactions,” in IEEE International Conference on Robotics and Automation (ICRA), 2023, pp. 7713–7719.
  55. M. Faessler, A. Franchi, and D. Scaramuzza, “Differential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high-speed trajectories,” IEEE Robotics and Automation Letters, vol. 3, no. 2, 2017.
  56. P. Forni, D. Jeltsema, and G. A. Lopes, “Port-Hamiltonian formulation of rigid-body attitude control,” IFAC-PapersOnLine, vol. 48, no. 13, 2015.
  57. R. Rashad, F. Califano, and S. Stramigioli, “Port-Hamiltonian passivity-based control on SE(3) of a fully actuated UAV for aerial physical interaction near-hovering,” IEEE Robotics and Automation Letters, vol. 4, no. 4, 2019.
  58. J. Delmerico and D. Scaramuzza, “A benchmark comparison of monocular visual-inertial odometry algorithms for flying robots,” in IEEE International Conference on Robotics and Automation (ICRA), 2018.
  59. S. A. S. Mohamed, M. Haghbayan, T. Westerlund, J. Heikkonen, H. Tenhunen, and J. Plosila, “A survey on odometry for autonomous navigation systems,” IEEE Access, vol. 7, 2019.
  60. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Kopf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala, “PyTorch: An Imperative Style, High-Performance Deep Learning Library,” in Advances in Neural Information Processing Systems 32, 2019, pp. 8024–8035.
  61. T. Lee, M. Leok, and N. H. McClamroch, “Geometric tracking control of a quadrotor UAV on SE(3),” in IEEE Conference on Decision and Control (CDC), 2010.
  62. J. Panerati, H. Zheng, S. Zhou, J. Xu, A. Prorok, and A. P. Schöllig, “Learning to fly: a pybullet gym environment to learn the control of multiple nano-quadcopters,” https://github.com/utiasDSL/gym-pybullet-drones, 2020.
  63. L. Meier, D. Honegger, and M. Pollefeys, “Px4: A node-based multithreaded open source robotics framework for deeply embedded platforms,” in IEEE International Conference on Robotics and Automation (ICRA), 2015, pp. 6235–6240.
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
Github Logo Streamline Icon: https://streamlinehq.com

GitHub

  1. Port-Hamiltonian Neural ODE Networks on Lie Groups For Robot Dynamics Learning and Control
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets

Youtube Logo Streamline Icon: https://streamlinehq.com

YouTube