Application of an improved SVM algorithm for wind speed forecasting

Huaqiang Zhang; Xinsheng Wang; Yinxiao Wu

doi:10.1007/978-3-642-19706-2_43

Application of an improved SVM algorithm for wind speed forecasting

Huaqiang Zhang^*
, Xinsheng Wang
, Yinxiao Wu

^*Corresponding author for this work

Harbin Institute of Technology Weihai

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Scopus citations

Abstract

An improved Support Vector Machine (SVM) algorithm is used to forecast wind in Doubly Fed Induction Generator (DFIG) wind power system without aerodromometer. The mathematical model is built after analyzing the principle of wind forecasting with Maximum Power Point Tracing (MPPT), and its kernel functions of SVM is selected. Compares the rapidity and accuracy of parameter optimization methods, we know that the Particle Swarm Optimization (PSO) method is better than the Cross Validation (CV) method. Finally, 3.6MW DFIG wind power system simulation model with wind speed forecasting is established. Simulation results show that the accuracy rate thought improved SVM forecasting algorithm can reach 98.667%, the DFIG system can operate at the MPPT. The whole performance has well robustness and rapidity.

Original language	English
Title of host publication	Future Intelligent Information Systems
Pages	333-340
Number of pages	8
Edition	VOL. 1
DOIs	https://doi.org/10.1007/978-3-642-19706-2_43
State	Published - 2011
Externally published	Yes
Event	2010 International Conference on Electrical and Electronics Engineering, ICEEE 2010 - Wuhan, China Duration: 4 Dec 2010 → 5 Dec 2010

Publication series

Name	Lecture Notes in Electrical Engineering
Number	VOL. 1
Volume	86 LNEE
ISSN (Print)	1876-1100
ISSN (Electronic)	1876-1119

Conference

Conference	2010 International Conference on Electrical and Electronics Engineering, ICEEE 2010
Country/Territory	China
City	Wuhan
Period	4/12/10 → 5/12/10

Keywords

Forecast
Maximum power point tracing (MPPT)
Particle swarm optimization
Support vector machine
Wind power generation

Access to Document

10.1007/978-3-642-19706-2_43

Cite this

@inproceedings{fcbb5dbdd63a434ab890d1e817ce78c6,

title = "Application of an improved SVM algorithm for wind speed forecasting",

abstract = "An improved Support Vector Machine (SVM) algorithm is used to forecast wind in Doubly Fed Induction Generator (DFIG) wind power system without aerodromometer. The mathematical model is built after analyzing the principle of wind forecasting with Maximum Power Point Tracing (MPPT), and its kernel functions of SVM is selected. Compares the rapidity and accuracy of parameter optimization methods, we know that the Particle Swarm Optimization (PSO) method is better than the Cross Validation (CV) method. Finally, 3.6MW DFIG wind power system simulation model with wind speed forecasting is established. Simulation results show that the accuracy rate thought improved SVM forecasting algorithm can reach 98.667\%, the DFIG system can operate at the MPPT. The whole performance has well robustness and rapidity.",

keywords = "Forecast, Maximum power point tracing (MPPT), Particle swarm optimization, Support vector machine, Wind power generation",

author = "Huaqiang Zhang and Xinsheng Wang and Yinxiao Wu",

year = "2011",

doi = "10.1007/978-3-642-19706-2\_43",

language = "英语",

isbn = "9783642197055",

series = "Lecture Notes in Electrical Engineering",

number = "VOL. 1",

pages = "333--340",

booktitle = "Future Intelligent Information Systems",

edition = "VOL. 1",

note = "2010 International Conference on Electrical and Electronics Engineering, ICEEE 2010 ; Conference date: 04-12-2010 Through 05-12-2010",

}

Zhang, H , Wang, X & Wu, Y 2011, Application of an improved SVM algorithm for wind speed forecasting. in Future Intelligent Information Systems. VOL. 1 edn, Lecture Notes in Electrical Engineering, no. VOL. 1, vol. 86 LNEE, pp. 333-340, 2010 International Conference on Electrical and Electronics Engineering, ICEEE 2010, Wuhan, China, 4/12/10. https://doi.org/10.1007/978-3-642-19706-2_43

TY - GEN

T1 - Application of an improved SVM algorithm for wind speed forecasting

AU - Zhang, Huaqiang

AU - Wang, Xinsheng

AU - Wu, Yinxiao

PY - 2011

Y1 - 2011

N2 - An improved Support Vector Machine (SVM) algorithm is used to forecast wind in Doubly Fed Induction Generator (DFIG) wind power system without aerodromometer. The mathematical model is built after analyzing the principle of wind forecasting with Maximum Power Point Tracing (MPPT), and its kernel functions of SVM is selected. Compares the rapidity and accuracy of parameter optimization methods, we know that the Particle Swarm Optimization (PSO) method is better than the Cross Validation (CV) method. Finally, 3.6MW DFIG wind power system simulation model with wind speed forecasting is established. Simulation results show that the accuracy rate thought improved SVM forecasting algorithm can reach 98.667%, the DFIG system can operate at the MPPT. The whole performance has well robustness and rapidity.

AB - An improved Support Vector Machine (SVM) algorithm is used to forecast wind in Doubly Fed Induction Generator (DFIG) wind power system without aerodromometer. The mathematical model is built after analyzing the principle of wind forecasting with Maximum Power Point Tracing (MPPT), and its kernel functions of SVM is selected. Compares the rapidity and accuracy of parameter optimization methods, we know that the Particle Swarm Optimization (PSO) method is better than the Cross Validation (CV) method. Finally, 3.6MW DFIG wind power system simulation model with wind speed forecasting is established. Simulation results show that the accuracy rate thought improved SVM forecasting algorithm can reach 98.667%, the DFIG system can operate at the MPPT. The whole performance has well robustness and rapidity.

KW - Forecast

KW - Maximum power point tracing (MPPT)

KW - Particle swarm optimization

KW - Support vector machine

KW - Wind power generation

UR - https://www.scopus.com/pages/publications/79956275488

U2 - 10.1007/978-3-642-19706-2_43

DO - 10.1007/978-3-642-19706-2_43

M3 - 会议稿件

AN - SCOPUS:79956275488

SN - 9783642197055

T3 - Lecture Notes in Electrical Engineering

SP - 333

EP - 340

BT - Future Intelligent Information Systems

T2 - 2010 International Conference on Electrical and Electronics Engineering, ICEEE 2010

Y2 - 4 December 2010 through 5 December 2010

ER -

Application of an improved SVM algorithm for wind speed forecasting

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this