Characterizing Time Series Data Diversity for Wind Forecasting

Cong Feng, Erol Chartan, Brian Hodge, Jie Zhang

Research output: Contribution to conferencePaperpeer-review

10 Scopus Citations

Abstract

Wind forecasting plays an important role in integrating variable and uncertain wind power into the power grid. Various forecasting models have been developed to improve the forecasting accuracy. However, it is challenging to accurately compare the true forecasting performances from different methods and forecasters due to the lack of diversity in forecasting test datasets. This paper proposes a time series characteristic analysis approach to visualize and quantify wind time series diversity. The developed method first calculates six time series characteristic indices from various perspectives. Then the principal component analysis is performed to reduce the data dimension while preserving the important information. The diversity of the time series dataset is visualized by the geometric distribution of the newly constructed principal component space. The volume of the 3-dimensional (3D) convex polytope (or the length of ID number axis, or the area of the 2D convex polygon) is used to quantify the time series data diversity. The method is tested with five datasets with various degrees of diversity.

Original languageAmerican English
Pages113-119
Number of pages7
DOIs
StatePublished - 5 Dec 2017
Event4th IEEE/ACM International Conference on Big Data Computing, Applications and Technologies, BDCAT 2017 - Austin, United States
Duration: 5 Dec 20178 Dec 2017

Conference

Conference4th IEEE/ACM International Conference on Big Data Computing, Applications and Technologies, BDCAT 2017
Country/TerritoryUnited States
CityAustin
Period5/12/178/12/17

Bibliographical note

See NREL/CP-5D00-71412 for preprint

NREL Publication Number

  • NREL/CP-5D00-70235

Keywords

  • Big data alization
  • Data diversity
  • Machine learning
  • Time series analysis
  • Wind forecasting

Fingerprint

Dive into the research topics of 'Characterizing Time Series Data Diversity for Wind Forecasting'. Together they form a unique fingerprint.

Cite this