DocumentCode :
2972657
Title :
Research on Control Method of Complex Object Based on Human Behavior Data
Author :
Zhong Bing-xiang ; Li Tai-fu ; Wang De-biao ; Su Ying-ying
Author_Institution :
Coll. of Electron. Inf. Eng., Chongqing Univ. of Sci. & Technol., Chongqing, China
fYear :
2010
fDate :
25-27 June 2010
Firstpage :
5486
Lastpage :
5489
Abstract :
In order to simulate human behavior to achieve intelligent control, in this paper a mathematical modeling method is presented based on Kernel Principal Component Analysis (KPCA) and Support Sector Machine. Sample data from the input space are mapped to high-dimensional feature space by non-linear transformation, then their features are extracted by PCA to decrease dimension of input vector and then the modeling based on support vector regression(SVR) are completed. Through the simulation of skilled people to drive a bike, mathematical model of complex system has reconstructed and intelligent control of complex objects is realized. By simulation and testing on experimental systems, the results are proved that the system model has high precision and good control effect.
Keywords :
control system analysis; intelligent control; large-scale systems; principal component analysis; regression analysis; support vector machines; bike; complex object; complex system; control method; high-dimensional feature space; human behavior data; intelligent control; kernel principal component analysis; mathematical modeling; nonlinear transformation; support sector machine; support vector regression; Artificial neural networks; Feature extraction; Fuzzy systems; Kernel; Mathematical model; Principal component analysis; Support vector machines; KPCA; SVR; behavior simulation; complex object; feature extraction;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Electrical and Control Engineering (ICECE), 2010 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-6880-5
Type :
conf
DOI :
10.1109/iCECE.2010.1333
Filename :
5629507
Link To Document :
بازگشت