Stochastic unit commitment with significant wind penetration using novel scenario generation and reduction

Wind energy has become a remarkable source of energy over the last decade. However, the uncertainty of wind prediction results in a great number of problems for the power systems operators in the day-ahead unit commitment. In order to maintain the power system reliable, the uncertainty of forecasting should be considered. In the field of data-mining, there are a number of techniques to model the uncertainties. This paper presents a novel solution for stochastic unit commitment with modeling the uncertainties of wind prediction. The uncertainty of demand has been considered as well. To do so, a new method for scenario generation and reduction is developed and solved with a modified shuffled frog leaping algorithm (SFLA) named adaptive SFLA (ASFLA). The case study shows the capability of the proposed method and a comparison indicated the benefits of wind energy employment in a system.