Fault diagnosis is critical to ensure the safety and reliable operation of rotating machinery. Most methods used in fault diagnosis of rotating machinery extract a few feature values from vibration signals for fault diagnosis, which is a dimensionality reduction from the original signal and may omit some important fault messages in the original signal. Thus, a novel diagnosis method is proposed involving the use of a convolutional neural network (CNN) to directly classify the continuous wavelet transform scalogram (CWTS), which is a time-frequency domain transform of the original signal and can contain most of the information of the vibration signals. In this method, CWTS is formed by discomposing vibration signals of rotating machinery in different scales using wavelet transform. Then the CNN is trained to diagnose faults, with CWTS as the input. A series of experiments is conducted on the rotor experiment platform using this method. The results indicate that the proposed method can diagnose the faults accurately. To verify the universality of this method, the trained CNN was also used to perform fault diagnosis for another piece of rotor equipment, and a good result was achieved.

Accurate fault information is critical for optimal scheduling of production activities, improving system reliability, and reducing operation and maintenance cost. In recent years, many fault diagnosis methods for rolling element bearings have been developed based on deep learning. Most of them are totally data-driven and do not consider the domain knowledge that has been used in fault diagnosis for years. Meanwhile, operating conditions such as rotating speed and load that have great influence on vibration signals are also ignored. It may cause a decrease in accuracy when bearing type or operating condition changes. To address these problems, this paper proposes a rolling element bearing fault diagnosis and localization approach based on multi-task convolutional neural network (CNN) with information fusion. In the proposed approach, domain knowledge, operating conditions, and vibration signals are fused into a 3-dimensional input that can be processed well by CNN. Then a multi-task CNN with dynamic training rates is constructed to simultaneously accomplish two tasks, fault diagnosis and localization. Experimental results on two rolling element bearing testbeds with different bearing types and operating conditions are presented and compared with existing state-of-the-art approaches to demonstrate the effectiveness and accuracy of the proposed approach.

Deep learning methods have been introduced for fault diagnosis of rotating machinery. Most methods have good performance when processing bearing data at a certain rotating speed. However, most rotating machinery in industrial practice has variable working speed. When processing the bearing data with variable rotating speed, the existing methods have low accuracies, or need complex parameter adjustments. To solve this problem, a fault diagnosis method based on continuous wavelet transform scalogram (CWTS) and Pythagorean spatial pyramid pooling convolutional neural network (PSPP-CNN) is proposed in this paper. In this method, continuous wavelet transform is used to decompose vibration signals into CWTSs with different scale ranges according to the rotating speed. By adding a PSPP layer, CNN can process CWTSs in different sizes. Then the fault diagnosis of variable rotating speed bearing can be carried out by a single CNN model without complex parameter adjustment. Compared with a spatial pyramid pooling (SPP) layer that has been used in CNN, a PSPP layer locates as front layer of CNN. Thus, the features obtained by PSPP layer can be delivered to convolutional layers for further feature extraction. According to experiment results, this method has higher diagnosis accuracy for variable rotating speed bearing than other methods. In addition, the PSPP-CNN model trained by data at some rotating speeds can be used to diagnose bearing fault at full working speed.

Recently, the machine-vision-based blades surface damage detection technique has received great attention for its low cost, easy operation, and lack of a need for prior knowledge. The rapid progress of deep learning has contributed to the promotion of this technology with automatic feature extraction, a broader scope of application, and stronger expansibility. An image recognition method of wind turbine blade defects using attention-based MobileNetv1-YOLOv4 and transfer learning is proposed in this paper. The backbone convolution neural network of YOLOv4 is replaced by the lightweight MobileNetv1 for feature extraction to reduce complexity and computation. Attention-based feature refinement with three distinctive modules, SENet, ECANet, and CBAM, is introduced to realize adaptive feature optimization. To solve the problem of slow network convergence and low detection accuracy caused by insufficient data, a two-stage transfer learning approach is introduced to fine-tune the pre-trained network. Comparative experiments verify the efficacy of the proposed model, with higher detection accuracy but a significantly faster response speed and less computational complexity, compared with other state-of-the-art networks by using images of the wind turbine blades taken by an unmanned aerial vehicle (UAV). A sensitivity study is also conducted to present the effects of different training dataset sizes on the model performance.