Air Conditioning Load Prediction Based on DE-SVM Algorithm

Author

Chen, Zhonghai ; Sun, Yong ; Yang, Guoli ; Wu, Tengfei ; Li, Guizhu ; Xin, Longbiao

Author_Institution

Dept. of Urban Constr., Hebei Inst. of Archit. & Civil Eng., Zhangjiakou, China

fYear

2010

fDate

2-4 April 2010

Firstpage

276

Lastpage

279

Abstract

Based on SVM (Support Vector Machine) theory, and the model to predict air conditioning load was established. In order to optimize the behavior of SVM, the DE (Differential Evolution) algorithm was introduced into classic SVM. The DE-SVM model is applied to a real example. The comparisons between the predicted results of the three models-GA (Genetic Algorithm) model, ACO (Ant Colony Optimization) and POS (Partial Swarm Optimization) show the precision of the DE-SVM algorithm is high with the maximum relative error being 2.52%.

Keywords

air conditioning; differential equations; evolutionary computation; genetic algorithms; load forecasting; particle swarm optimisation; power engineering computing; support vector machines; ACO; DE-SVM algorithm; POS; air conditioning load prediction; ant colony optimization; differential evolution algorithm; genetic algorithm; maximum relative error; partial swarm optimization; support vector machine; Air conditioning; Ant colony optimization; Civil engineering; Information technology; Kernel; Linear regression; Mathematical model; Prediction algorithms; Predictive models; Support vector machines; Air Conditioning load; DE-SVM; Prediction;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Information Technology and Security Informatics (IITSI), 2010 Third International Symposium on

Conference_Location

Jinggangshan

Print_ISBN

978-1-4244-6730-3

Electronic_ISBN

978-1-4244-6743-3

Type

conf

DOI

10.1109/IITSI.2010.41

Filename

5453625