Forecasting photovoltaic production

A simple and intuitive Flask application for forecasting photovoltaic production from 1 to 14 days ahead. It uses the pvlib library to model the photovoltaic system with few parameters and open-meteo API to obtain the forecasts.

Project configuration

• Clone the project from the repository:

  git clone https://github.com/Giudice7/pv_forecast

• Open the terminal and move to the project folder.

  cd pv_forecast

• Check the python version installed:

  python --version

  In this project, Python 3.11 was used, so it is recommended to use the same version to avoid compilation problems. If you have a different version, download the 3.11 from the following link.

• Create a virtual environment (venv) in the project folder:

  python -m venv venv

• Activate the virtual environment:

  venv/Scripts/activate

• Install dependencies:

  pip install -r requirements.txt

• Run the application:

  python app.py

• Open the browser and go to the following address:

  http://127.0.0.1:5000

Adapt to your PV system

• Open the file config.json and change the parameters of the photovoltaic system to match your system. The parameters are:
  • latitude: Latitude of the location in decimal degrees.
  • longitude: Longitude of the location in decimal degrees.
  • forecasting_days: Number of days to forecast.
  • pdc0_inverter: Nominal power of the inverter in W.
  • array_list: The list of arrays that compose the whole PV system.

Each array has the following parameters:

• surface_tilt: The tilt angle of the panels.
• surface_azimuth: The azimuth angle of the panels.
• module_type: The module materials sandwich: can be one of polymer_glass or glass_glass.
• pdc0_module: The nominal power of each module.
• gamma_pdc_module: The loss coefficient due to temperature increase of the module (from - 0.002 to -0.005 typically)
• module_per_string: The number of module per each string.
• strings: The number of strings in the array.

Running in docker

• Build the image:

  docker build -t pv_forecast .

• Run the container:

  docker run -p 5000:5000 pv_forecast

• Open the browser and go to the following address:

  http://localhost:5000

References

1. pvlib-python
2. open-meteo API