Skip to Main Content
This paper proposes a method combining support vector data description (SVDD) and fussy control to deal with the gluing procedure. First, Vibration signals are used as data of fault diagnosis, kernel principal component analysis (KPCA) is employed for feature extraction of the normal and fault examples, and SVDD algorithm as classifier is used for fault diagnosis. Second, fuzzy control rules are designed for the normal signals in order to keep the quantity of the glue with the variation of the fiber. Experiments show that SVDD algorithm is practical and efficient and the control process based on fuzzy control is stable and reliable.