Gas peak demand forecasting in Tunisia using machine learning

This project aimed to forecast natural gas peak demand of Mseken city (Sousse, Tunisia) in long-term, mid-term and short-term. Different variables were taken into account: gas flow, number of customers, holidays, temperature,...

Main Contribution

The first paper to study natural gas demand forecasting in Tunisia using machine learning techniques A part of the work done was published in this article:

Install dependencies

Install libraries listed in the notebooks.

Data

Data for this research cannot be shared.

Description of each file

-EDA.ipynb : exploratory data analysis.
-Forecasting_temperature_and_clients.ipynb : forecasting of the number of clients and temperature -Forecasting_until_2024.ipynb : forecasting of the gas peak demand from 2021 to 2024 -Pics_cleaning.ipynb : data cleaning -The rest of the notebooks represent a plethora of statistical, ML and DL methods implemented to forecast gas peak demand over different horizons.

📬 Contact

If you have questions or encounter issues, please open an issue or contact us at sami.benbrahim@mail.concordia.ca.

📖 Citation

If you use this code or find our work helpful, please consider citing:

@INPROCEEDINGS{9765941,
  author={Brahim, Sami Ben and Slimane, Mohamed},
  booktitle={2022 5th International Conference on Advanced Systems and Emergent Technologies (IC_ASET)}, 
  title={Long-term natural gas peak demand forecasting in Tunisia Using machine learning}, 
  year={2022},
  volume={},
  number={},
  pages={222-227},
  keywords={Support vector machines;Energy resolution;Demand forecasting;Machine learning;Learning (artificial intelligence);Predictive models;Maintenance engineering;natural gas;peak flow;long-term forecasting;machine learning},
  doi={10.1109/IC_ASET53395.2022.9765941}}