Multi-task learning applied to wind farms production forecasting

• Grant number: 20/03913-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2020
• End date: May 31, 2021
• Field of knowledge: Engineering - Electrical Engineering
• Principal Investigator: Fernando José von Zuben
• Grantee: Gabriel Ribeiro Lencione
• Host Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

The widespread implementation of wind farms integrated to electric power systems all over the world, specially in Brazil, makes the ability to estimate the future power production through wind speed forecasting very important. Once there are historical time series of wind speed intensity associated to multiple wind farms, the multiple tasks of time series prediction should be solved simultaneously. This project of scientific initiation proposes the adoption of consolidated Multi-Task Learning techniques, allowing information exchange among forecasting tasks that demonstrate to be related. The information sharing must conduct to improvement of prediction performance when compared to treating the forecasting tasks independently. Besides the practical relevance of this research and the use of real study cases, the student will have the opportunity to get involved in a large field of machine learning concepts and to get in touch with a research group which has a consistent history of contributions in the field of machine learning applied to climate variables. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

LENCIONE, GABRIEL R.; VON ZUBEN, FERNANDO J.; IEEE. Wind Speed Forecasting via Multi-task Learning. 2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), v. N/A, p. 8-pg., 2021-01-01. (20/03913-0)