An Improved Colony Clustering to Short-Term Forecasting in the Power System

In this paper, an Improved Ant Colony Clustering (IACC) based on Ant Colony Algorithm is presented. In IACC, each load data was represented by an ant, and the merits of IACC were parallel search optimum and the dynamic method to adjust the evaporation coefficient, which can raise the forecast accuracy. IACC used the weighted Euclidean distance, and the residual pheromone quantity was calculated by the radius of the cluster and the weighted Euclidean distance, in the end, IACC ameliorated the warp error of the cluster canter. What´s more, IACC was more sensitive to weather condition and date than the Ant Colony Optimization Algorithm (ACOA). The comparison between IACC and ACOA shows that IACC increased the short- term load forecasting (STLF) accuracy, and IACC is more exquisite to the similarity of load curve profile.