Classification and summarization of solar irradiance and power forecasting methods: A thorough review

Abstract

Solar forecasting is of great importance for ensuring safe and stable operation of power system with increased solar power integration, thus numerous models have been presented and reviewed to predict solar irradiance and power in the past decade. Nevertheless, few studies take into account temporal and spatial resolution along with specific characteristics of models. Therefore, this paper aims to make a comprehensive and systematic review for further solve these problems. Firstly, five classifications and seven pre-processing methods of solar forecasting data are systematically reviewed, which are significant in improving forecasting accuracy. Then, various methods utilized in solar irradiance and power forecasting are thoroughly summarized and discussed, in which 128 algorithms are elaborated in tables in the light of input variables, temporal resolution, spatial resolution, forecast variables, metrics, and characteristics for a more fair and comprehensive comparison. Moreover, they are categorized into four groups, namely, statistical, physical, hybrid, and others with relevant application conditions and features. Meanwhile, six categories along with 30 evaluation criteria are summarized to clarify major purposes/applicability of different methods. The prominent merit of this work is that a total of seven perspectives and trends for further research in solar forecasting, which aims to help readers more effectively utilize these approaches for future in-depth researches.