By Topic

Time Series Prediction Based on Improved Arithmetic of Support Vector Regression

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Chao Zhang ; Dept. of Mech. Eng., North China Electr. Power Univ., Baoding ; Pu Han

Traditional method of mathematical modeling, such as statistical theory and artificial neural network, usually gets a non-linear model for complicated system of turbine. Based on the need of modeling for turbine shafting vibration, a new improved arithmetic of support vector regression (svr), which is named as smooth support vector regression (SSVR), is imported and used for time series analysis and prediction. SSVR tries to establish a linear model in high-dimension feature space, so the model built by SSVR is easily to reflect the implicit mechanism of shafting vibration data set. Simulations of sinc series data and actual turbine vibration data indicate that SSVR advances the training capability of standard SVR method, and is very fit for the time series of small sample size. Prediction experiment of turbine shafting vibration shows that the SSVR arithmetic has higher training and predicting precision, and gets better generalization ability in the mean time, which is obviously superior to neural network method.

Published in:

2008 Fourth International Conference on Natural Computation  (Volume:2 )

Date of Conference:

18-20 Oct. 2008