Fault diagnosis is important in monitoring the entire lifecycle of mechanical equipment. And Feature selection can explain the relationships between different fault types and reduce the complexity of the fault diagnosis model. However, due to the correlation among the fault features, a single method is unable to effectively determine the importance of the features. In order to enhance the analytical ability of the feature selection for features and thereby improve the universality of the method, this paper proposes a new feature analysis method – the multi-factor dual analysis method. This method is designed to conduct in-depth analysis of the fault data and utilize statistical methods to determine the optimal ranking of feature importance. By testing on different fault data sets, it has been proved that this method can effectively reduce the feature dimension while maintaining the classification accuracy.
Yang et al. (Mon,) studied this question.