Fuzzy control system

Control (management)

Engineering

Fuzzy inference system

Statistics

Support vector machine

Adaptive neuro fuzzy inference system

Inference system

Particle swarm optimization

Computer science

Mechanical engineering

Wind speed

Physics

Regression

Mathematics

Geography

Fuzzy logic

Inference

Artificial intelligence

Turbine

Wind power

Meteorology

Control theory (sociology)

Artificial neural network

Neuro-fuzzy

Machine learning

Electrical engineering

Sustainable Energy Technologies and Assessments

In this study, three models of machine learning algorithms are implemented to predict wind speed, wind direction and output power of a wind turbine. The first model is multilayer feed-forward neural network (MLFFNN) that is trained with different data training algorithms. The second model is support vector regression with a radial basis function (SVR-RBF). The third model is adaptive neuro-fuzzy inference system (ANFIS) that is optimized with a partial swarm optimization algorithm (ANFIS-PSO). Temperature, pressure, relative humidity and local time are considered as input variables of the models. A large set of wind speed and wind direction data measured at 5-min, 10-min, 30-min and 1-h intervals are utilized to accurately predict wind speed and its direction for Bushehr. Energy and exergy analysis of wind energy for a wind turbine (E-44, 900 kW) is done. Also, the developed models are used to predict the output power of the wind turbine. Comparison of the statistical indices for the predicted and actual data indicate that the SVR-RBF model outperforms the MLFFNN and ANFIS-PSO models. Also, the current energy and exergy analysis presents an average of 32% energy efficiency and approximately 25% exergy efficiency of the wind turbine in the study region.

Prediction of wind speed and wind direction using artificial neural network, support vector regression and adaptive neuro-fuzzy inference system