A novel hybrid metaheuristic optimization method to estimate medium-term output power for horizontal axis wind turbine


EKINCI F., Demirdelen T., AKSU I. O., AYGÜL K., ESENBOGA B., BİLGİLİ M.

PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART A-JOURNAL OF POWER AND ENERGY, vol.233, no.5, pp.646-658, 2019 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 233 Issue: 5
  • Publication Date: 2019
  • Doi Number: 10.1177/0957650918821040
  • Journal Name: PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART A-JOURNAL OF POWER AND ENERGY
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.646-658
  • Keywords: Engineering optimization, wind power prediction, artificial neural network, metaheuristic optimization, firefly algorithms, NEURAL-NETWORK, SPEED, ENSEMBLE, DECOMPOSITION, ALGORITHM, SELECTION, MODELS, RELM
  • Çukurova University Affiliated: Yes

Abstract

The increasing damage caused by fossil fuels has made it a necessity for new and clean energy sources. In recent years, the use of wind energy from renewable energy sources has increased, which is a new and clean energy source. Wind energy is everywhere in nature. The wind speed changes depending on time. Thus, the wind power is unstable. In order to keep this disadvantage at a minimum level, future power estimation studies have been carried out. In these studies, different methods and algorithms are applied to estimate short and medium term in wind power. In this study, artificial neural network, particle swarm optimization and firefly algorithm (FA) as a new method are used for the first time in predicting wind power. As input data, temperature, wind speed and rotor speed the data recorded in the SCADA in wind turbines are used to predict medium-term wind speed and also wind power. Each method is compared in detail and their performances are revealed.