Fault diagnosis with fault gradation using neural network group
GU Xiao-dong · Systems engineering and electronics · 2009
In order to resolve the shortage of the neural network fault diagnosis and improve the precision and veracity,a new fault diagnosis approach with fault gradation using neural network group consisting of 3 sub neural networks is proposed.The faults different grades are given according to the occurrence frequencies of different faults.The higher the fault grade,the larger the number of the used sub neural networks is.Experimental results show that the diagnosis correctness rate of the faults with the highest grade is 100%,and the diagnosis correctness rates of the other faults with lower grades,are about 95%.The proposed approach is of better performance.