Probability Forecast of Photo Voltaic Power with an Adaptive Discretization Algorithm in Higher-Order Markov Chain

The higher-order Markov chain (HMC) model is most common photovoltaic (PV) output probability prediction method. The HMC model requires to determine the number of discretization states subjectively, a technique to optimize the selection of the state number, however, is missing. To address this issue, an adaptive discretization method based on information entropy theory is proposed in this paper, which considers the PV output volatility and enhances prediction accuracy and uncertainty capturing ability. Experimental results using real-world PV output datasets validates the superiority of the proposed method over traditional HMC models by achieving higher prediction accuracy and better capture the uncertainty.