Cloud tracking using clusters of feature points for accurate solar irradiance nowcasting

详细信息    查看全文

• 作者：Hsu-Yung Cheng ^{chengsy@csie.ncu.edu.tw}
• 关键词：Cloud tracking ; Feature point ; Clustering ; Irradiance nowcasting ; Ramp-down event
• 刊名：Renewable Energy
• 出版年：2017
• 出版时间：April 2017
• 年：2017
• 卷：104
• 期：Complete
• 页码：281-289
• 全文大小：1909 K
• 卷排序：104

文摘

Cloud tracking is achieved using clusters of feature points. An enhanced clustering algorithm is proposed to prevent over-clustering or under-clustering. Obscuration events are predicted with features of clustering and tracking information. A graph based model is designed to prediction ramp-down events. Detection results are applied in irradiance nowcasting frameworks to improve accuracy.