Decision support for renewable energy


Icare has developed an algorithm capable of predicting the consumption and production of renewable energy. This solution is being applied in several projects aimed at optimising electricity production, both for industry and for private individuals.
The digital solution developed by Icare provides a forecast of consumption and production of renewable solar energy, whether for a single customer installation or for a geographical area. It can be done for days or for the next hour. A Machine Learning algorithm uses a wide range of data: consumption and production history, weather archives, forecasts from MeteoSwiss. The model integrates many other parameters, including holidays, weekends, etc.
Advantages: the algorithm works 24 hours a day, processes a large volume of data and is faster and more accurate than humans. Forecasting is substantially improved. For industry, but also for private individuals, being able to obtain a forecast of their electricity consumption or production provides a valuable decision-making aid. Depending on the data transmitted by the predictive model, is it better to consume the energy produced, or to store it? In all cases, the aim is to make maximum use of the electricity produced on site, avoiding having to send it to the grid. The time horizon for a reliable prediction is currently limited to 72 hours.
This decision support for consumption and production forecasts for the following day is currently used on a daily basis by several energy companies in Valais and outside the canton.