Fitting a mixture of von Mises distributions in order to model data on wind direction in Peninsular Malaysia

Research output: Contribution to journalArticle

41 Citations (Scopus)

Abstract

A statistical distribution for describing wind direction provides information about the wind regime at a particular location. In addition, this information complements knowledge of wind speed, which allows researchers to draw some conclusions about the energy potential of wind and aids the development of efficient wind energy generation. This study focuses on modeling the frequency distribution of wind direction, including some characteristics of wind regime that cannot be represented by a unimodal distribution. To identify the most suitable model, a finite mixture of von Mises distributions were fitted to the average hourly wind direction data for nine wind stations located in Peninsular Malaysia. The data used were from the years 2000 to 2009. The suitability of each mixture distribution was judged based on the R2 coefficient and the histogram plot with a density line. The results showed that the finite mixture of the von Mises distribution with H number of components was the best distribution to describe the wind direction distributions in Malaysia. In addition, the circular density plots of the suitable model clearly showed the most prominent directions of wind blows than the other directions.

Original languageEnglish
Pages (from-to)94-102
Number of pages9
JournalEnergy Conversion and Management
Volume72
DOIs
Publication statusPublished - 2013

Fingerprint

Potential energy
Wind power

Keywords

  • Circular density
  • Mixture of von Mises distribution
  • Statistical model
  • Wind direction
  • Wind regime

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Fuel Technology
  • Nuclear Energy and Engineering
  • Renewable Energy, Sustainability and the Environment

Cite this

@article{aa6b37f88ec2493ca56cc3b18b1b485a,
title = "Fitting a mixture of von Mises distributions in order to model data on wind direction in Peninsular Malaysia",
abstract = "A statistical distribution for describing wind direction provides information about the wind regime at a particular location. In addition, this information complements knowledge of wind speed, which allows researchers to draw some conclusions about the energy potential of wind and aids the development of efficient wind energy generation. This study focuses on modeling the frequency distribution of wind direction, including some characteristics of wind regime that cannot be represented by a unimodal distribution. To identify the most suitable model, a finite mixture of von Mises distributions were fitted to the average hourly wind direction data for nine wind stations located in Peninsular Malaysia. The data used were from the years 2000 to 2009. The suitability of each mixture distribution was judged based on the R2 coefficient and the histogram plot with a density line. The results showed that the finite mixture of the von Mises distribution with H number of components was the best distribution to describe the wind direction distributions in Malaysia. In addition, the circular density plots of the suitable model clearly showed the most prominent directions of wind blows than the other directions.",
keywords = "Circular density, Mixture of von Mises distribution, Statistical model, Wind direction, Wind regime",
author = "Nurulkamal Masseran and Razali, {Ahmad Mahir} and Kamarulzaman Ibrahim and Latif, {Mohd Talib}",
year = "2013",
doi = "10.1016/j.enconman.2012.11.025",
language = "English",
volume = "72",
pages = "94--102",
journal = "Energy Conversion and Management",
issn = "0196-8904",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Fitting a mixture of von Mises distributions in order to model data on wind direction in Peninsular Malaysia

AU - Masseran, Nurulkamal

AU - Razali, Ahmad Mahir

AU - Ibrahim, Kamarulzaman

AU - Latif, Mohd Talib

PY - 2013

Y1 - 2013

N2 - A statistical distribution for describing wind direction provides information about the wind regime at a particular location. In addition, this information complements knowledge of wind speed, which allows researchers to draw some conclusions about the energy potential of wind and aids the development of efficient wind energy generation. This study focuses on modeling the frequency distribution of wind direction, including some characteristics of wind regime that cannot be represented by a unimodal distribution. To identify the most suitable model, a finite mixture of von Mises distributions were fitted to the average hourly wind direction data for nine wind stations located in Peninsular Malaysia. The data used were from the years 2000 to 2009. The suitability of each mixture distribution was judged based on the R2 coefficient and the histogram plot with a density line. The results showed that the finite mixture of the von Mises distribution with H number of components was the best distribution to describe the wind direction distributions in Malaysia. In addition, the circular density plots of the suitable model clearly showed the most prominent directions of wind blows than the other directions.

AB - A statistical distribution for describing wind direction provides information about the wind regime at a particular location. In addition, this information complements knowledge of wind speed, which allows researchers to draw some conclusions about the energy potential of wind and aids the development of efficient wind energy generation. This study focuses on modeling the frequency distribution of wind direction, including some characteristics of wind regime that cannot be represented by a unimodal distribution. To identify the most suitable model, a finite mixture of von Mises distributions were fitted to the average hourly wind direction data for nine wind stations located in Peninsular Malaysia. The data used were from the years 2000 to 2009. The suitability of each mixture distribution was judged based on the R2 coefficient and the histogram plot with a density line. The results showed that the finite mixture of the von Mises distribution with H number of components was the best distribution to describe the wind direction distributions in Malaysia. In addition, the circular density plots of the suitable model clearly showed the most prominent directions of wind blows than the other directions.

KW - Circular density

KW - Mixture of von Mises distribution

KW - Statistical model

KW - Wind direction

KW - Wind regime

UR - http://www.scopus.com/inward/record.url?scp=84878950498&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84878950498&partnerID=8YFLogxK

U2 - 10.1016/j.enconman.2012.11.025

DO - 10.1016/j.enconman.2012.11.025

M3 - Article

VL - 72

SP - 94

EP - 102

JO - Energy Conversion and Management

JF - Energy Conversion and Management

SN - 0196-8904

ER -