Statistical Approaches for Wind Resource Assessment


Abstract

We describe and evaluate a set of computational intelligence techniques for long-term wind resource assessment. Short term sensor measurements at a potential wind farm site are correlated with pub-licly available wind data sources in close spatial proximity in order to extrapolate long-term predictions for the site. This general approach to assessment is called \MCP": Measure, Correlate and Predict. Our tech-niques are based upon statistical inference. They aim to address accurately correlating inexpensive but noisy, short term measurements at the site. Each technique relies upon estimating the joint distribution of wind speeds at the site and the publicly available neighbouring sources. For a site at the Boston Museum of Science when the availability of site data varies between 3, 6 and 8 months, we find that copula modeling is robust to data availability and consistently best overall.
Authors: Kalyan Veeramachaneni, Xiang Ye, and Una-May O'Reilly

Source: and Full Article:
http://groups.csail.mit.edu/EVO-DesignOpt/groupWebSite/uploads/Site/WRE-Chapter.pdf

0 yorum: