Fault diagnosis of rolling linear guides based on wavelet packet decomposition and support vector machines

Rolling linear guideways are standard equipment in the machining industry, and their failures can directly affect productivity. This paper presents a fault diagnosis method based on wavelet packet decomposition (WPD) and support vector machines (SVM) for monitoring rolling linear guideways, focusing on detecting faults in the guideways. Vibration signals from the guideway are collected using acceleration sensors and decomposed using WPD to extract the energy in each frequency band as features. By training these features, the detection of rail faults can be achieved. The proposed fault diagnosis method was tested on a testbed, and an online monitoring and fault diagnosis system for rolling linear guideways was designed. Experimental results show that the system achieves a prediction accuracy of 95% and that the method is feasible and effective for moving linear guideway fault diagnosis.