Reinforcement learning policies using elapsed time since odor detection and exponentially filtered local wind direction outperform cast-and-surge in simulated turbulent plumes with mild mean wind and show optimal performance at intermediate memory times in isotropic turbulence.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
physics.flu-dyn 2verdicts
UNVERDICTED 2representative citing papers
Heterogeneous agent policies outperform homogeneous ones in collective olfactory search within 3-D turbulent odor fields generated by Navier-Stokes DNS.
citing papers explorer
-
Smart strategies to navigate turbulent odor plumes reorienting to local wind
Reinforcement learning policies using elapsed time since odor detection and exponentially filtered local wind direction outperform cast-and-surge in simulated turbulent plumes with mild mean wind and show optimal performance at intermediate memory times in isotropic turbulence.
-
Policy heterogeneity improves collective olfactory search in 3-D turbulence
Heterogeneous agent policies outperform homogeneous ones in collective olfactory search within 3-D turbulent odor fields generated by Navier-Stokes DNS.