A kind of K — Nearest neighbor fault diagnosis method based on MIV data transformation

K-Nearest Neighbor (KNN) is a commonly used fault diagnosis method, which is based on Euclidean distance between samples to carry out fault diagnosis. The differences between the variables have a direct effect on the Euclidean distance, which affects the KNN fault diagnosis effect. After the dimensional normalization, there are also some problems such as the decrease of variable diversity, and the geometry is evenly distributed. In order to solve the above problems, this paper introduces the concept of Mean Impact Value (MIV), and establishes a method of evaluating the contribution of components to BP neural network. Based on the contribution of each component, the original data is transformed and the new KNN method based on MIV is established. Firstly, the sample data is normalized. Secondly, the MIV value of each characteristic variable after data normalization is calculated by BP neural network. Furthermore, carry out fault diagnosis based on the fault diagnosis model created. Finally, the effectiveness of the proposed method is verified by the simulation test of UCI standard data set.