Use of Artificial Neural Networks for Predicting the Offshore Wind Turbine Power Curve Using Various Training Algorithms in Newfoundland

Low predictive accuracy of energy output is one of the weakest points in wind power generation. Power curve of a wind turbine provides assistance in energy assessment, warranty formulations, and performance monitoring of the wind turbines. Wind turbines are being installed in offshore diverse climatic conditions causing substantial departure of these power curves from the warranted values. As industrial organizations being managed by enterprise-wide systems, a software like solution for the prediction of wind farm power output is desirable. The proposed wind farm performance prediction models should be able to forecast the amount of energy produced on different time scales such as 10 min, 1 hour and a day. Such prediction models would transform a wind farm into a wind power plant. Artificial Neural Network is such model that has been used in this study to accurately forecast the power output of a proposed offshore wind farm.

Faculty Supervisor:

Jianming James Yang

Student:

Partner:

Springboard Atlantic Inc.

Discipline:

Engineering

Sector:

Green/Alternative Energy; Clean Technology; Artificial Intelligence

University:

Memorial University of Newfoundland

Program: