Characterizing Time Series Data Diversity for Wind Forecasting: Preprint

Brian Hodge, Cong Feng, Jie Zhang

Research output: Contribution to conferencePaper


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 1D 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
Number of pages10
StatePublished - 2018
EventIEEE/ACM International Conference on Big Data Computing, Applications, and Technologies - Austin, Texas
Duration: 5 Dec 20178 Dec 2017


ConferenceIEEE/ACM International Conference on Big Data Computing, Applications, and Technologies
CityAustin, Texas

Bibliographical note

See NREL/CP-5D00-70235 for paper as published in proceedings

NREL Publication Number

  • NREL/CP-5D00-71412


  • big data visualization
  • data diversity
  • machine learning
  • time series analysis
  • wind forecasting


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

Cite this