A data-driven approach to the performance analysis of wind turbines is presented. Turbine performance is captured with a power curve. The power curves are constructed using historical wind turbine data. Three power curve models are developed, one by the least squares method and the other by the...

A data-driven approach for the optimization of a heating, ventilation, and air conditioning (HVAC) system in an office building is presented. A neural network (NN) algorithm is used to build a predictive model since it outperformed five other algorithms investigated in this paper. The NN-derived...

Different models for monitoring wind farm power output are considered. Data mining and evolutionary computation are integrated for building the models for prediction and monitoring. Different models using wind speed as input to predict the total power output of a wind farm are compared and...