Piecewise Aggregate Approximation and quantile regression for wind speed analysis

Ronaldo R. B. de Aquino, Helen Barboza da Silva, Jonata C. de Albuquerque, Manuel Herrera, Aida A. Ferreira, Alcides Codeceira Neto

Research output: Chapter or section in a book/report/conference proceedingChapter in a published conference proceeding


The high cost of energy production, coupled with the advantages of wind power as renewable and widely available source of energy, has led several countries to establish incentives to
regulate and promote wind power generation. This work proposes the implementation and comparison of two time series analysis methods: Piecewise Aggregate Approximation (PAA) and a PAA plus quantile regression process. The aim is to estimate the minimum amount of extreme cut-in and cut-out events of the wind speed in the power generation process. Brazil has an enormous wind power potential. The diversification of its energy matrix is becoming a necessary challenge nowadays in the commitment of using renewable energy sources. The performance of the two PAA based proposals is tested for to the wind farms of the south and
northeast regions. These locations belong to Brazil’s regions of different geographic and wind characteristics. This endows to also check the proposals robustness under divergent scenarios. The results indicate that the PAA/QR method performed better than the PAA method because it identified a greater amount of extreme values.
Original languageEnglish
Title of host publicationAnnual Conf. on Computational Science & Computational Intelligence
Publication statusAcceptance date - 28 Oct 2017
EventInternational Conference on Computational Science and Computational Intelligence - Las Vegas, USA United States
Duration: 14 Dec 201716 Dec 2017


ConferenceInternational Conference on Computational Science and Computational Intelligence
Abbreviated titleCSCI'17
Country/TerritoryUSA United States
CityLas Vegas
Internet address


Dive into the research topics of 'Piecewise Aggregate Approximation and quantile regression for wind speed analysis'. Together they form a unique fingerprint.

Cite this