Content area
In this work we evaluate the impact of considering a stochastic approach on the day-ahead basis Unit Commitment. Comparisons between stochastic and deterministic Unit Commitment solutions are provided.
The Unit Commitment model consists in the minimization of the total operation costs considering units’ technical constraints like ramping rates and minimum up and down time. Load shedding and wind power spilling is acceptable, but at inflated operational costs.
The generation of Unit Commitment solution is guaranteed by DEEPSO, which is a hybrid DE-EA-PSO algorithm, where DE stands for Differential Evolution, EA for Evolutionary Algorithms and PSO for Particle Swarm Optimization.
The evaluation process consists in the calculation of the optimal economic dispatch and in verifying the fulfillment of the considered constraints. For the calculation of the optimal economic dispatch an algorithm based on the Benders Decomposition, namely on the Dual Dynamic Programming, was developed. If possible, the constraints added to the dispatch problem by the Benders Decomposition algorithm will provide a feasible and optimal dispatch solution.
Two approaches were considered on the construction of stochastic solutions. Either the top 5 more probable wind power output scenarios are used, or a set of extreme scenarios are considered instead.
Data related to wind power outputs from two different operational days is considered on the analysis. Stochastic and deterministic solutions are compared based on the actual measured wind power output at the operational day. Through a technique capable of finding representative wind power scenarios and their probabilities we were able to analyze in a more detailed process the expected final operational costs. Also, we expose the probability that the system operator has on the operational costs being under/above certain value.
Results show that the stochastic approach leads to more robust Unit Commitment solutions than the deterministic one. The method of using the top 5 more probable scenarios on the search for the stochastic solution proved to produce preferable results.