Clustering of Typical Wind Power Scenarios Based on K-Means Clustering Algorithm and Improved Artificial Bee Colony Algorithm

Abstract

An improved artificial bee colony algorithm (IABC) was proposed to solve the problems of the k-means clustering (KMC) algorithm, such as poor global search ability, sensitive selection of initial cluster center, randomness of initialization, precocity and slow convergence of the original artificial bee colony (ABC) algorithm. To improve the efficiency of iterative optimization process, a fitness function adapted to KMC algorithm and a position updating formula based on global guidance were constructed. By comparing the improved artificial bee colony algorithm with the original artificial bee colony algorithm and particle swarm optimization algorithm, it is confirmed that IABC algorithm converges speed block and overcomes the shortcoming of the original algorithm which is easy to fall into local optimal solution. IABC algorithm is combined with KMC to get better clustering effect, and the algorithm is used to select typical wind power output scenarios, which plays an important role in actual production.