Nonlinear Estimation of Important Variables using Neural Network in Wind Farms

Utilising additional variables, which are normally not measured, could bring significant improvements to condition monitoring and control of a wind turbine and farm. However, incorporating additional sensors to measure such variables could increase the cost significantly. As a solution, instead of equipping every turbine in the wind farm with an expensive sensor to measure such a variable, it is proposed that only one turbine be equipped with a sensor and the neighbouring turbines with an estimator that essentially replaces the sensor; that is, each estimator would subsequently estimate what the sensor would measure. Each estimator is constructed based on Neural Network, and as a result, the cost could be significantly reduced. Note that the only turbine equipped with a sensor is used to train the NN. This work presents the results of a preliminary study to examine the feasibility of the proposed approach.