Multi-Feature Fusion and RF-Based Fault Diagnosis of Ball Mill Rolling Bearings

Abstract

Due to the complex working conditions of the metallurgical industry, it is difficult to obtain high-quality fault features from a single signal, and the diagnosis effect is poor. Aiming at the problem that the direct fusion of current and vibration signals cannot reflect the advantages of the two types of signals in different frequency bands and the complementary information between each other, which affects the diagnostic performance, a multi-feature complementary fusion fault diagnosis method based on vibration and current signals is proposed. The high-frequency coefficient features of the vibration signal and the current signal are fused by the maximum absolute value rule to form complementary features that reflect the high-frequency band features; the low-frequency coefficient features of the vibration signal and the current signal are fused by sparse representation (SR) to form complementary features that reflect the low-frequency band features. By defining a feature matrix composed of multiple features to fuse the full-frequency band features, the global feature representation capability is enhanced. The recursive feature elimination method is used to eliminate the redundant features after fusion to improve the classification accuracy, and the bearing fault status is classified in combination with random forest (RF). The experimental results show that the proposed method is more accurate than the diagnosis results based on vibration signals and current signals.