文章基本信息

标题：Modified Support Vector Regression Model For Very Short-Term Solar Irradiance Forecasting
本地全文：下载
作者：Gokul Sidarth Thirunavukkarasu ; Mehdi Seyedmahmoudian ; Elmira Jamei 等
期刊名称：Webology
印刷版ISSN：1735-188X
出版年度：2022
卷号：19
期号：3
页码：2027-2038
语种：English
出版社：University of Tehran
摘要：One of the most reliable and prominent renewable energy resource which addresses the global energy demand is the Solar photovoltaics (PV) systems. Usage of the PV system substantially reduces the carbon footprint of the energy generation process leading towards an environmentally friendly alternative. The substantial rise in the adaptation of the distributed energy resources likes solar PV and batteries into conventional electric power grids, has increased the complexity of the energy management problem. Many Researchers across the world are working towards developing accurate forecasting models to predict the solar PV system’s generation capacity in advance to effectively manage the supply and demand. In this research, a modified support vector regression (SVR) based solar irradiance forecasting model with a minute wise time horizon is proposed, developed, and tested using the data obtained from Bureau of Meteorology (BOM), Australia. The performance of the proposed approach is then validated using the benchmark statistical models like, the persistence algorithm, autoregression model, moving average model, a hybrid autoregressive moving average model and an autoregressive integrated moving average model. Metrics like root mean square error (RMSE), the mean absolute error (MAE) and mean bias error (MBE) are used to validate the performance of the proposed approach. Results indicated that the Modified SVR model reduces the forecasting error and the computational complexity substantially and outperforms the other conventional approaches. The increased performance of the forecasting method will assist in developing an efficient energy management system for future electricity grids.
关键词：Machine learning;PV power forecasting;Renewable energy resources;Support vector Regression