Optimal Generator Scheduling using Probabilistic Demand Scenarios

Abstract

We have developed a new method to determine an optimal generator scheduling using probabilistic demand scenarios. The demand scenarios are created statistically as a set of load curves by the past demand data. Covariance matrix is used to reconstruct the distribution of the demand scenarios. Unit commitment calculation is executed for each demand scenario and the operation cost is estimated. Optimal generator scheduling is selected by the expected value of the unit price with including the additional cost to compensate the demand variation. The risk of the scenarios is also estimated to select the optimal generation scenario. The advantage of this method is that the optimum generator scheduling can be calculated in the present load dispatching system since the existing scheduling algorism can be easily adopted with a minimum modification.