Review on Probabilistic Short-Term Power Forecast

Abstract

Wind power, photovoltaic (PV) generation and loads are influenced by lots of factors, which causes different levels of errors existed in their forecast results. Compared with traditional point forecast methods, probabilistic forecast methods can quantize uncertainties in the process of forecast, provide more comprehensive and valuable probability distribution information for power system operation and control. To obtain a clearer development venation, this paper makes a unified review of wind power, PV generation and load power probabilistic forecast methods. Firstly, this paper analyzes the causes of forecast uncertainties, then summarizes the merits, demerits and applicable conditions of some probabilistic forecast processes and the common used models. After that this paper collects the popular evaluation indexes and existing evaluation mechanism. At the end, the future development directions of wind power, PV generation and load power probabilistic forecast is prospected.