A journal article published in August has attempted to address the issue of indeterminacy in wave energy by creating a technique designed to predict the energy in waves.
Published in the journal Ocean Engineering, the paper — Short-term forecasting of the wave energy flux: Analogues, random forests, and physics-based models — analyzed the performance of four separate models used in forecasting wave energy flux.
Marine energy has a great potential for generating the renewable energy necessary for a world that, for the most part, sees a need to step away from a reliance upon fossil fuels. The water is always there, the moon does its thing, and we can rely on waves to be creating the energy we need in a way we can’t always predict and rely on wind and sun.
It’s also easier to predict than predicting wind and sun, which allows utilities to be better informed about the possible influx of energy coming into their grids.
Knowing in advance when an increase in electricity is on the way is an important part of integrating any renewable energy into existing electricity grids.
The EOLO group, based out of the University of the Basque Country, developed a variety of models for predicting the amount of wave energy that is caused in the Bay of Biscay by using a technique called random forests. Gabriel Ibarra, of the EOLO group, explains:
“Random forests (RF) is an algorithm developed in recent years in the field of machine learning. The basis of RFs are the so-called ‘regression trees’, in which the input variables are regarded as roots and the output ones, the leaves. Hence the name ‘tree’. Random forest is a development of the regression trees which uses many trees (over a thousand, as a general rule) rather than just one, thus forming a forest”
The University of the Basque Country goes into a bit more detail about the ways in which these models acquire their information and the further focus of the EOLO group.