AI Science Tracker
EN

Deep reinforcement learning with spatial and temporal awareness for active boundary control of buoyancy-driven convection

Not stated in the abstract reviewed · 29 Oct 2026 · physics

Discovery

Deep reinforcement learning with spatial and temporal awareness for active boundary control of buoyancy-driven convection

Source check: Source checked. Not independent scientific confirmation.

Finding: Deep reinforcement-learning policies with spatial and temporal awareness controlled simulated buoyancy-driven convection. Reported simulations changed heat and salt transport, including reduced Nusselt number in a Rayleigh–Bénard case and improved heat transfer with lower salinity variance in a double-diffusive case. These are computational results, not experiments.

Technology: CNN policies with GRU memory trained using TD3/MADDPG

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
CNN policies with GRU memory trained using TD3/MADDPG
Test and result
Controlled simulations compared learned policies with baselines using flow and transport metrics across reported convection settings.
Limitation / still unknown
Simulation-only result; no physical experiment or operational control demonstration is reported.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://openaccess.city.ac.uk/id/eprint/38576/1/1-s2.0-S0142727X26005163-main.pdf

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse physics discoveries.