DocumentCode :
2286885
Title :
The study about feature selection of analog circuit fault diagnosis based on annealing genetic hybrid algorithm
Author :
Haijun, Lin ; Yao, Fu ; Zhicheng, Xu ; Xuhui, Zhang
Author_Institution :
Harbin University of Science and Technology, Harbin China
fYear :
2012
fDate :
18-21 Sept. 2012
Firstpage :
1
Lastpage :
4
Abstract :
This paper forms the annealing genetic hybrid algorithm, which is about the genetic algorithm and simulated annealing algorithm improved and integrated. For the problem about feature selection of nonlinear analog circuit fault diagnosis based on Volterra kernel, using annealing genetic hybrid algorithm to research, put forward annealing genetic intelligent selection method of circuit fault diagnosis feature. Experiments show that this method has realized effective selection of analog circuit diagnostic features and improved the prediction accuracy of fault diagnosis system based on BP neural network.
Keywords :
analogue circuits; backpropagation; circuit analysis computing; fault diagnosis; genetic algorithms; nonlinear network analysis; simulated annealing; BP neural network; Volterra kernel; annealing genetic intelligent selection method; nonlinear analog circuit fault diagnosis feature selection; simulated annealing genetic hybrid algorithm; Algorithm design and analysis; Circuit faults; Fault diagnosis; Genetic algorithms; Genetics; Optimization; Sociology; Analog circuits; Annealing genetic algorithms; Fault diagnosis; Feature selection;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Strategic Technology (IFOST), 2012 7th International Forum on
Conference_Location :
Tomsk
Print_ISBN :
978-1-4673-1772-6
Type :
conf
DOI :
10.1109/IFOST.2012.6357817
Filename :
6357817
Link To Document :
بازگشت