Photovoltaic Power Forecasting: Models and Methods

Abstract

The integration of PV power brings many economic and environmental benefits. However, high penetration of PV power brings many challenges for power system operations and planning, mainly due to its uncertain and intermittent characteristics. The output power of PV system is subjected to ramping, since its dependence on solar irradiance and other meteorological factors changes frequently. One of the possible solutions to balance these challenges is power output forecasting. This paper presents a comprehensive and systematic overview of existing research works on PV power forecasting. Major factors that influence the output of PV power generation and PV power forecasting are discussed. The research progress of statistical methods for photovoltaic power forecasting in recent years is also investigated. In addition, model selection strategy and assessment indicators are analyzed. The effect of model input, the potential benefits of hybrid methods, and the effect of parameter selection in PV power forecasting are further explored.