From Probabilistic PV Power Forecasting to Stochastic Programing: a Method for Microgrid Planning Under Uncertainty

Recently, probabilistic PV forecasting emerged at a good level of maturity, questioning its role and potential in microgrid planning. Forecasts can take various forms but this paper focuses firstly on a procedure to obtain empirical distribution from quantiles. Secondly, correlation structure of studied variables is explored using copula theory. After addressing scenario Monte-Carlo generation and reduction methods, the interest of stochastic programming using created trajectories is discussed.