Wind Power Forecasting Error Distributions and Probabilistic Load Dispatch

Among renewable power sources, wind energy is the most promising technology; however, the inter-temporal uncertainty of this source makes impossible its massive integration. Forecasting of wind generation is a key factor for the economical operation of the power system. Thus, the error related to this process is typically modeled by means of a determined probability distribution to be later incorporated to the unit scheduling and load dispatch optimization procedures. In this paper, wind power forecasting error has been modeled by using Weibull and Levy alpha-stable probability distributions and incorporated to the economic dispatch problem in order to probabilistically describe power production and generating cost. The proposed methodology is illustrated by analyzing a case study composed by 13 conventional generators; the obtained results are compared with Monte Carlo Simulation approach for evaluating and testing the capabilities of the proposed model, observing reasonable accuracy on the estimated results.