Title :
Research of fault diagnosis method of analog circuit based on improved support vector machines
Author :
Li, Hua ; Yin, Bin ; Li, Nan ; Guo, Jianhua
Author_Institution :
Sch. of Inf. Sci. & Eng., Northeastern Univ., Shenyang, China
Abstract :
This paper propose improved support vector machine algorithm. The algorithm includes preprocessing the sample training set, improvement of the binary tree classification algorithm and incremental sample learning algorithm. Considering the specific classification precision requirements of analog circuit fault diagnosis, the three algorithms are integrated, and achieve good results. The simulation of analog circuit demonstrate that the improved algorithm has higher classification precision and faster diagnosis speed compared to traditional support vector machine algorithm.
Keywords :
Analog circuits; Artificial neural networks; Circuit faults; Classification algorithms; Dictionaries; Fault diagnosis; Fault tolerance; Neural networks; Support vector machine classification; Support vector machines; Analog Circuit; Fault Diagnosis; Improved Support Vector Machine;
Conference_Titel :
Industrial Mechatronics and Automation (ICIMA), 2010 2nd International Conference on
Conference_Location :
Wuhan, China
Print_ISBN :
978-1-4244-7653-4
DOI :
10.1109/ICINDMA.2010.5538189