Weather type partition method considering sequential features in photovoltaic forecasting. Issue 13 (13th December 2017)

• Record Type: Journal Article
• Authors: Li, Hao; Lu, Zongxiang; Qiao, Ying; Wang, Ningbo; Ding, Kun
• Is Part Of: Journal of engineering. Volume 2017:Issue 13(2017)
• Page Start: 1259
• Page End: 1263
• Publication Date: 2017-12-13
• Subjects: load forecasting -- photovoltaic power systems -- prediction theory -- self‐organising feature maps -- power engineering computing; self‐organising map neural network algorithm -- characteristic index system -- sequential difference sequence analysis -- PV data characteristics -- photovoltaic prediction method -- photovoltaic forecasting -- weather type partition method; Engineering -- Periodicals; Engineering; Electronic journals; Periodicals; 620.005
• DOI: 10.1049/joe.2017.0531
• Languages: English
• ISSNs: 2051-3305

Abstract

Abstract : The prediction method based on the partition of weather types has been widely used in photovoltaic prediction. Here, the temporal characteristics of PV data is studied, the physical meaning of the sequential difference sequences is analysed, and a characteristic index system including sequential characteristics index is proposed. The weather types are partitioned based on self‐organising map neural network algorithm. Case study shows that the partition method is helpful to improve the discrimination of weather types and to improve prediction accuracy.