Path integral control is extended to Gaussian belief space for partially observed systems, yielding necessary and sufficient matching conditions, an exact Cole-Hopf linearization via Feynman-Kac, and the MPPI-Belief algorithm that outperforms certainty-equivalent and particle-filter baselines on a导航
Online algorithms for POMDPs with continuous state, action, and observation spaces,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.SY 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Path Integral Control in Gaussian Belief Space for Partially Observed Systems
Path integral control is extended to Gaussian belief space for partially observed systems, yielding necessary and sufficient matching conditions, an exact Cole-Hopf linearization via Feynman-Kac, and the MPPI-Belief algorithm that outperforms certainty-equivalent and particle-filter baselines on a导航