A hidden Markov model representing the spatial and temporal correlation of multiple wind farms

To accommodate the increasing wind energy with stochastic nature becomes a major issue on power system reliability. This paper proposes a methodology to characterize the spatiotemporal correlation of multiple wind farms. First, a hierarchical clustering method based on self-organizing maps is adopted to categorize the similar output patterns of several wind farms into joint states. Then the hidden Markov model (HMM) is then designed to describe the temporal correlations among these joint states. Unlike the conventional Markov chain model, the accumulated wind power is taken into consideration. The proposed statistical modeling framework is compatible with the sequential power system reliability analysis. A case study on optimal sizing and location of fast-response regulation sources is presented.