PH-Dreamer: A Physics-Driven World Model via Port-Hamiltonian Generative Dynamics

TL;DR

This paper introduces PH-Dreamer, a physics-driven world model that incorporates port-Hamiltonian principles to improve physical consistency, efficiency, and fidelity in visual control tasks.

Contribution

It presents a unified port-Hamiltonian framework embedding physical priors, a kinematics-aware energy model, and an energy-guided Actor-Critic for enhanced control and simulation fidelity.

Findings

Achieves superior asymptotic returns on visual control benchmarks.

Reduces latent phase space volume by 4.18-8.41%.

Lowers energy consumption by up to 7.80%.

Abstract

World models built on recurrent state space architectures enable efficient latent imagination, yet remain physically unstructured, producing dynamics that violate conservation and dissipative principles. We introduce a unified Port-Hamiltonian framework that remedies this through three synergistic mechanisms. First, we embed implicit physical priors into recurrent transitions by modeling projected latent evolution as action controlled energy routing governed by flow and dissipation, biasing the projected PH phase space toward a more compact and physically structured representation. Second, we develop a kinematics aware energy world model that estimates the Hamiltonian and power balance from proprioceptive observations, providing an explicit physical signal for thermodynamic reasoning. Third, leveraging these energy gradients, we establish an energy guided Actor-Critic that uses Lagrangian multipliers to regularize policy optimization toward lower energy and smoother control. Across visual control benchmarks, this paradigm not only attains superior asymptotic returns but also elevates internal simulator fidelity by establishing a tighter, lower variance alignment between imagined and real rewards, all while reducing latent phase space volume by 4.18-8.41%, energy consumption by up to 7.80%, and mean squared jerk by up to 9.38%.