Neural Network Approach to Pitch Angle Control in Wind Energy Conversion Systems for Increased Power Generation

Abstract

Presented in this study is an artificial intelligence approach to pitch angle control in wind turbines for the enhancement of the power generation efficiency of wind energy conversion systems. A two-input neural network model was developed and trained using backward propagation technique to adjust the pitch angle of the turbine in response to the speed of turbine generator and the rate of change of the speed. Ten-year real-life data on the wind speeds of a study location was used to validate the approach. It was found that the method performs well in controlling the mechanical power developed by the turbine above the turbine’s rated wind speed, and with fast processing time. 44.44%improvement was achieved in the mechanical power developed by the turbine. The control approach is thus recommended for the effective management of wind energy conversion systems towards enhancing availability and reliability of electric power supply.