cx_rl_gym
Base class wrapping gym-functions for reinforcement learning to interface with CLIPS over cx_rl_interfaces.
README
CXRLGym: ROS 2 Reinforcement Learning Environment
The CXRLGym module provides a ROS 2-integrated Gymnasium environment for reinforcement learning (RL) with multiple robots.
It acts as a bridge between the CLIPS-Executive and standard RL frameworks, allowing agents to perceive and act in a symbolic robot environment using familiar Gym interfaces.
Overview
CXRLGym extends gymnasium.Env to provide a ROS-driven reinforcement learning interface.
It automatically constructs its observation and action spaces by querying ROS 2 services and actions.
Key Features
Dynamic observation and action spaces generated from ROS 2 services
Multi-robot management and coordination via reinforcement learning
Compatible with Gym-based RL libraries (e.g., Stable Baselines3)
ROS 2 action interfaces for reset, robot allocation, and action execution