mppt_td3

An implementation oT Twin delayed DDPG (TD3) (https://spinningup.openai.com/en/latest/algorithms/td3.html) for maximum power point traking of a photovoltaic cell.

Requirements:

• Tensorflow 1
• numpy
• python 3.5
• Gym Open AI
• Gym environment 'mppt-v0', 'mppt-v1'

Source: mppt-v0

Clone the mmppt-v0 gym environment from https://github.com/loavila/mppt-gym and install following the instructions.

It is recommendable installing in a python virtual environment (https://rukbottoland.com/blog/tutorial-de-python-virtualenv/)

Environment descriptions

• mmpt-v0 is a standard environment of a photovoltaic system (put link to the model)

• mmpt_shaded-v0 is an environment of a photovoltaic system with partial shading (put link to the model)

How to run:

In a console run:

python main.py

Testing:

python simul_test2.py