A HYBRID ARTIFICIAL NEURAL NETWORK-DYNAMIC PROGRAMMING APPROACH TO UNIT COMMITMENT

In this paper, a hybrid dynamic programming-artificial neural network algorithm (DP-ANN) is studied. The proposed two-step process uses an artificial neural network (ANN) to generate a pre-schedule according to the input load profile. Then a dynamic search is performed at those stages where the commitment states of some of the units are not certain. The experimental results indicate that the proposed algorithm can significantly reduce the execution time of the traditional dynamic programming approach without degrading the quality of the generation schedule.