DC 4: Use of machine learning tools for estimating EGs performance.

The aim of this project is to investigate the possibility to use machine learning
techniques to predict the behaviour of EGs in terms of heat exchange and help in the optimisation of their functioning.

1. Collect monitoring and numerical data on different EG types and different functioning scenarios (heating-cooling modes, different external and climatic conditions).
2. Define input and output parameters adapted for the evaluation of the heat exchange of EGs, and design a model based on artificial neural networks. Train models and check their performance on available data.
3. Use the models to predict the performance of EGs under different future scenarios (change in climatic conditions, temperature, activation modes, etc).