Fault diagnosis of diaphragm pump check valve based on impulse and cyclostationary analysis

Extracting sensitive features from vibration signals is still a challenge for intelligent fault diagnosis. Although the fault diagnosis methods have achieved great success in the past decade. However, most of the methods are not able to combine domain knowledge effectively, which leads to inadequate ability fault diagnosis in the actual working environment. To address this issue, we propose a novel fault diagnosis method for the diaphragm pump check valve based on impulse and cyclostationary analysis. Firstly, the wavelet packet transform is used to decompose each intercepted signal to enhance the hidden specific signal structure in the transform domain; then, the impulse and cyclostationary features of each frequency band are extracted separately to capture the inherent properties of the check valve in different working states; finally, the extracted sample set is used as input to the classifier for pattern recognition. The results show that the proposed method can accurately identify the fault classes of the check valve.