Study of vibration fault diagnosis for hydro-turbine generating unit based on improved principal component analysis model
Wang Ruilia · Journal of Hydroelectric Engineering · 2014
The paper describes an improved principal component analysis model for hydro-turbine generating unit characterized by complicated causes of vibration faults and many fault parameters. This model uses the membership in fuzzy theory for normalization and standardization of the numerical values in the initial value matrix and selects all the principal components to integrate the selection standard and reduce information loss. For the principal components, their weights are specified by the theory of entropy utility value to represent the amount of information dimension of primal variable contained within them, and to meet the strict requirements by their volume and regularity, grey relational degree is used to calculate the weights for a comprehensive evaluation. A case study of a hydro-turbine generating unit indicates that the method is useful for vibration fault diagnosis involved with fault parameters of different physical quantities.