f-krause / uio-aiei-project-1

UiO Project 1 for AI for Energy Informatics: Deep Reinforcement Learning

Assignment 1 - IN5460

Felix Krause, Johannes Spieß, Sina Henning

Mandatory assignment 1 for Artificial Intelligence for Energy Informatics.

Assignment | Deadline 17.10.2023 kl 23:59

Setup

Environment setup

With conda

conda create --name ai4ei python=3.11
conda activate ai4ei
conda install -c conda-forge jupyter numpy pandas seaborn matplotlib
pip install gym

With venv

python -m venv venv

Activate env in bash:

source venv/bin/activate

Activate env in PowerShell (Windows)

C:\Code\in5460-mex1\venv\Scripts\Activate.ps1

Install packages in venv

pip install -r requirements.txt

Add virtual environment to kernel

python -m ipykernel install --user --name=venv

Provide Datasets

Put the following datasets into a folder "/data" in this repo

• SolarIrradiance.csv
• rate_consumption_charge.csv/electricity price
• WindSpeed.csv

Put the "BASE" profiles in a folder "/data/residential_load_data_base"

• Residential load profiles: Check out data source here

Run training

Open the notebook "notebooks/1_training.ipynb".

Analysis of results

Open the notebook "notebooks/2_result_viz.ipynb".

Conventions

• The main unit used for energy is kWh
• The used time frame is an hour

RL software

• pymgrid is discontinued, we looked at its successor python-microgrid
• python-microgrid can be used "to generate and simulate a large number of microgrids"; according to their data section, they als use OpenEI load data and PV datasets ("Data in pymgrid are based on TMY3 (data based on representative weather)", this might be EnergyPlus)
• easygrid: a simple version based on OpenAI gym, maybe we should use this to get started
• example tutorial an example project that is very similar to ours
• gym environments tutorial on how to create a RL environment
• stable baselines3 documentary of stable baselines3