Release Time:2022-06-26  Hits:

Date of Publication: 2008-01-01

Journal: 控制与决策

Issue: 3

Page Number: 302-305,309

ISSN: 1001-0920

Abstract: A least squares support vector machine (LS-SVM) based intelligent prediction method is proposed which uses iterative learning to deal with the varying elevator traffic. The future traffic demand is forecasted dynamically to find the varying regular pattern. A new principle of traffic pattern recognition based on the net increment and the intensity of gradient change of the traffic demand is presented. By constructing filter function of the traffic difference, the principal features of the varying traffic demand are extracted and the traffic pattern recognition is conducted to obtain the principal traffic pattern critical points. Simulation experiments show the effectiveness of the proposed method.

Note: 新增回溯数据