A practical method of unit commitment considering wind power

Abstract

In this paper, a dispersed probability distribution model is built based on statistics of wind power forecasting error. Thus, the probabilistic reserve constraints in the analytic expression can be introduced into the traditional unit commitment (UC) model by Gauss function fitting of the distributions of different generations considering wind power. Otherwise, in order to solve the UC problem with these reserve constraints efficiently, this paper presents an improved approach of modifying Lagrangian multipliers based on the concept of average cost in the unit decommitment (UD) method. In addition, the search range and economic indices of units are also improved. A 26-unit system is analyzed to exhibit the effectiveness of the probabilistic reserve constraints and the proposed UD method.