Entries for Net Load Forecasting Prize

Using Climate-smart solar irradiance predictions for day-ahead net load forecasting

Using ML to produce accurate day ahead net load forecasting.

OpenSTEF is a fully automated, open-source software stack that predicts future load on the electricity grid using machine learning.

To Ensure reliable energy supply ,combine energy sources and energy storage and forecast load to promote sustainable energy grid.

Determinate forecast of the sky darkening by the probabilities analysis of data able to calculate solar irradiation per day/hour.

This innovation combines probabilistic weather forecast information with self classifying artificial intelligence to predict load shapes.

Long term prediction using sequence models

We use data-driven techniques to segregate load and generation from the perceived net load at the retail node with tuned local dynamics

A probabilistic deep learning model will be developed to perform probabilistic net load forecasts using the provided challenge data.

We'll work on uncertainty quantification for machine learning models.

We propose a method that combines data-driven and physics-informed approaches to best estimate net load based on historical data.

We will develop an ensemble learning approach to stabilize the prediction accuracy and thus increase the algorithm robustness.