An Intelligent Fault Diagnosis Method in the Case of Rotating Speed Fluctuations

摘要

Effective fault diagnosis method has long been a hot topic in the field of prognosis and health management of rotary machinery. This paper investigates an effective deep learning method known as sparse filtering, which is used to extract features from fault signal directly. And then, the supervised learning method softmax regression is applied to classify the fault types. The training samples are the vibration signals under a certain rotational speed and the test samples are in different rotational speeds. The key parameters of the model are optimized and analyzed through orthogonal experiments and single factor experiment. The diagnosis results show that the sparse filtering model has strong robustness for rotating machinery fault diagnosis in the case of rotating speed fluctuations.