A Fault Diagnosis Method for Rotating Machinery Under Variable Speed Condition Based on Infrared Thermography

Rotating machinery always works under variable speed and heavy load, resulting in difficulty in fault diagnosis. To avoid the influence caused by variable speed and noise, a novel fault diagnosis framework based on infrared thermography (IRT) is proposed in this paper. First, the IRT technique is introduced to acquire thermal images. Second, Bag-of-visual-word (BoVW) is used to extract visual features from the thermal images. In the end, the extracted visual features are taken into Softmax regression classifier to recognize the fault types of rotating machinery. The effectiveness of the proposed method is validated using the experimental data. Results show that the performance of proposed method is superior to the vibration based method in identifying ten health conditions of rotating machinery under variable speed condition.