safe-autonomous-systems/ma-sac-RBC2D-hard-v0

SAC on RBC2D-hard-v0 (FluidGym)

This repository is part of the FluidGym benchmark results. It contains trained Stable Baselines3 agents for the specialized RBC2D-hard-v0 environment.

Evaluation Results

Global Performance (Aggregated across 5 seeds)

Mean Reward: -0.69 ± 0.27

Per-Seed Statistics

Run	Mean Reward	Std Dev
Seed 0	-0.31	1.37
Seed 1	-0.46	3.00
Seed 2	-1.04	2.89
Seed 3	-0.82	2.59
Seed 4	-0.84	3.04

About FluidGym

FluidGym is a benchmark for reinforcement learning in active flow control.

Usage

Each seed is contained in its own subdirectory. You can load a model using:

from stable_baselines3 import SAC
model = SAC.load("0/ckpt_latest.zip")

Important: The models were trained using fluidgym==0.0.2. In order to use them with newer versions of FluidGym, you need to wrap the environment with a FlattenObservation wrapper as shown below:

import fluidgym
from fluidgym.wrappers import FlattenObservation
from stable_baselines3 import SAC

env = fluidgym.make("RBC2D-hard-v0")
env = FlattenObservation(env)
model = SAC.load("path_to_model/ckpt_latest.zip")

obs, info = env.reset(seed=42)

action, _ = model.predict(obs, deterministic=True)
obs, reward, terminated, truncated, info = env.step(action)

References

• Plug-and-Play Benchmarking of Reinforcement Learning Algorithms for Large-Scale Flow Control
• FluidGym GitHub Repository