Adaptive rule generation for solar forecasting: Interfacing with a knowledge-base library

Abstract

This paper covers the development of an adaptive, interactive rule generation interface applied to the full scope of solar forecasting techniques, both current and forthcoming. The interface provides a user-friendly platform for detecting patterns and correlations between elements in a database of solar irradiance, weather and photovoltaic generation information. The database consists of 10 years of data obtained from the National Renewable Energy Laboratory (NREL) data acquisition systems and the Automated Surface Observing System (ASOS). This report discusses how such an interface can be used to improve existing forecasting algorithms and also be used to create new forecasting techniques.