CartPole RL Codegen

MATLAB/Simulink reinforcement-learning project for a DQN CartPole controller. The workflow trains a DQN agent in MATLAB/Simulink, separates the trained policy for deployment, and generates deployable C code for embedded targets.

The repository highlights the deployment side of reinforcement learning: trained policy separation, controller-only Simulink modeling, and generated C artifacts.

Contents

• cartpole_full_project_codegen_R2024b_FINAL.m: training, simulation, plotting, policy block generation, and codegen workflow
• cartpole_policy_codegen.slx: controller-only Simulink model for code generation
• cartpole_policy_codegen_ert_rtw/*.c and *.h: generated policy/controller C source

The repository keeps source and generated deployment code only. Generated caches, trained agent snapshots, plots, logs, local build metadata, and local machine paths are excluded.

Skills Shown

MATLAB | Simulink | DQN | Reinforcement Learning | Code Generation | Embedded Control