Estimation and Prediction in Energy Systems and Infrastructure

The course focuses on exploitation of data from renewable energy systems and greeninfrastructure. The data can be used for parametrization of grey-box models of these systems where physical laws are combined with data to yield models of appropriate complexity for various uses in renewable energy systems management and their corresponding variablesprediction. The first part of the course will deal with estimation of parameters of grey-box models of renewable energy systems from data. Students will be made aware of bias-variance trade-off in data-based modelling. The second part focuses on machine-learning-based assessment of prediction models of key variables in renewable energy systems and infrastructure. Here the importance of correct data preprocessing and analysis will be stressed before using some readily available machine learning tools. Suitable structures of prediction models for different types of data will be explained.