Study on the Method for Fault Sample Selection Based on Fuzzy Clustering
A new method of fault sample selection is presented. First, the mapping relationship is established between the system components and system functions attributes according to the similarity of the system components and system functions attributes, the components are clustered to the class by fuzzy clustering method; then the fault sample is selected in the class according to relation of the faults propagation, the selected fault sample is validated by using the function adequacy and test adequacy. Finally From experimental results, the conclusion can be drown that the cost of this method is lower and the fault detection ratio is high. Compare with the other method the new method has certain advantages.
Helen Zhang, Gang Shen and David Jin
H. X. Wang et al., "Study on the Method for Fault Sample Selection Based on Fuzzy Clustering", Advanced Materials Research, Vols. 219-220, pp. 492-495, 2011