arXiv:2606.31137v1 Announce Type: new Abstract: This paper proposes a Bayesian filtering-based approach for learning the dynamics of a physical system from partial, noisy measurements. We model the system dynamics using a Lagrangian mechanics formulation. As in Lagrangian neural networks (LNNs), we parameterize the kinetic and potential energies with neural networks. The unknown external forces in the Lagrangian formulation are modeled as white Gaussian noise. The corresponding Euler--Lagrange equations then yield a continuous-time stochastic state-space model (SSM) that describes the system d
The paper leverages recent advancements in neural networks and Bayesian filtering to address the challenge of learning complex physical system dynamics from noisy data, reflecting ongoing AI research trends.
This research provides a more robust and efficient method for modeling dynamic physical systems, crucial for applications in robotics, control, and broader engineering, especially with imperfect sensor data.
The ability to learn Lagrangian dynamics more accurately from noisy, partial measurements improves the fidelity and reliability of simulations and autonomous control systems, reducing reliance on perfectly clean data.
- · Robotics sector
- · Control systems engineers
- · AI/ML researchers
- · Autonomous vehicle developers
- · Systems heavily reliant on perfectly clean sensor data
- · Manual calibration processes for complex systems
Improved performance and reliability of physical AI systems due to better foundational dynamics learning.
Faster development and deployment of advanced autonomous systems, as they can adapt to real-world complexities more effectively.
Enhanced AI capabilities contributing to breakthroughs in areas requiring precise physical interaction, such as advanced manufacturing or human-robot collaboration.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG