Cuckoo search optimization for short term wind energy forecasting

Carlos Eduardo M. Barbosa,Germano C. Vasconcelos
DOI: https://doi.org/10.1109/CEC.2016.7744002
2016-07-01
Abstract:Energy produced by wind farms is already responsible for a relevant portion of world's renewable energy and is increasingly becoming more important in modern society. As energy produced by wind farms is highly dependent on wind speed, it is not generated with regular output values requiring good forecasting systems to ensure good electricity supply. Cuckoo Search is a method recently introduced as a promising optimization algorithm for parameter estimation in some aplications and in this work it is investigated in wind energy forecasting. Cuckoo search is compared to other popular nature inspired techniques to predict the output values of wind parks in Texas and Montana, both in the USA. Results showed, with statistical significance, Cuckoo Search (CS) and Self Adaptive Cuckoo Search (SACS) not only presented higher prediction performance when compared to PSO, ACO and a hybrid of both but also converged significantly faster than the other algorithms.
What problem does this paper attempt to address?