Rooftop solar photovoltaic power forecasting using characteristic generation profiles

Abstract

Rooftop solar photovoltaic generation systems are often subject to location and site-specific factors that affect the shape of the daily generation profile. We propose a novel approach to rooftop solar photovoltaic power forecasting that applies digital filtering to recent historical generation data to determine a site-specific "Characteristic Generation Profile". This profile can subsequently be used on its own, or in combination with any available exogenous variables, to improve the solar power forecasting process. The approach is simple to implement, uniquely customised for every system it is applied to, and computationally efficient.