Master Thesis: Residential Energy Consumption Optimization through Reinforcement Learning

Welcome to the repository dedicated to my Master's thesis focusing on Reinforcement Learning (RL) applications in household energy optimization.

Contents

• nyiso_hourly_prices.csv: Contains historical price data crucial for modeling and simulating energy market environments.
• requirements.txt: PIP package information essential for setting up the required Python environment.
• SmartHomeGymEnv_v2_deployment.py: Gymnasium environment definition essential for training RL algorithms tailored for energy market scenarios.
• train_agent_on_environment.ipynb: Notebook for training RL agents in various environmental contexts.
• test_agent_deployment.ipynb: Testing pre-trained RL agents' performance on specific days within the energy market.
• LICENCE.txt: The Licence for this work
• Diploma_Thesis.pdf: The PDF of my Master thesis with the title "Residential Energy Consumption Optimization through Reinforcement Learning"

Setup Instructions

Before Running:

1. Download and open Anaconda Prompt (Download Anaconda).
2. Execute: conda create -n your_env_name python=3.10.12 to create a virtual Conda environment with the specified Python version.
3. Activate the virtual Conda environment: conda activate your_env_name.
4. Install required packages: pip install -r requirements.txt.
5. Run the Jupyter Notebooks to explore RL training and testing procedures.

This repository serves as a comprehensive guide and resource hub for understanding, experimenting with, and implementing RL models in the context of household energy optimization.

Feel free to explore the provided resources, engage in experiments, and contribute to the advancement of Reinforcement Learning methodologies within this wonderful field.

Acknowledgements

I would like to express my sincere gratitude to my supervisor, Professor George C. Polyzos, for his continuous and valuable guidance and support during the preparation of this thesis and in my studies in general, and to Professor Iordanis Koutsopoulos for his helpful comments and constructive criticism. I am also very grateful to PhD candidate Spiros Chadoulos for the technical guidance and knowledge he provided me and the countless conversations we had, helping me to complete this thesis successfully. This work was supported by the EMERGENT project which has received funding from the 2nd I-NERGY open call under Grant 101016508.