Development of an Industrial Forecasting Tool in Wind Power

The rapid growth of wind generation is introducing additional variability and uncertainty into power system operations and planning. These inherent characteristics of wind power have both technical and commercial implications for efficient planning and operation of power systems. As the penetration of wind power increases, the importance of accurate forecasting of this variable generation source over a number of time frames becomes more important. In this paper, the state-of-the-art wind power forecast technologies and developed commercial forecasting software have been investigated. Furthermore, the contribution of this paper is to design an industrial tool for wind power forecasting, which can be utilized to train forecasting models, predict wind power, extract proper historical data, compute forecasting residuals, accelerate the forecasting speed with the modified Levenberg-Marquardt algorithm, and executing combination forecasts with Genetic Algorithm method.