As vital renewable energy devices, wind turbines suffer from gearbox failures due to harsh speed increasing operations. Therefore, gearbox fault diagnosis is crucial for wind turbine maintenance with reducing economic costs. However, obtaining faulty data is rather challenging, especially at the early fault stage. For this reason, a sparse isolation encoding forest (SIEF) is proposed aiming at both anomaly detection and novel fault discrimination for wind turbine gearboxes. In the present SIEF method, a sparse autoencoder (SAE) is first trained with only normal data to obtain an optimized and robust weight structure. Newly acquired data corresponding to faulty or healthy conditions are sent to this encoder for feature extraction by encoding to its low-dimensional space. All the data in low-dimensional space are fed to an isolation forest (IF) for anomaly detection and novel fault discrimination. In the addressed SIEF approach, only normal data are required to train the model for fault detection and further discrimination. It is consistent with the actual operations of the wind turbines, with much less dependence on the fault data for model training. The proposed method was evaluated by fault diagnosis tests on wind turbine gearboxes. Results show good performances of the proposal compared with peers.

Induction motors, the key equipment for rotating machinery, are prone to compound faults, such as a broken rotor bars and bearing defects. It is difficult to extract fault features and identify faults from a single signal because multiple fault features overlap and interfere with each other in a compound fault. Since current signals and vibration signals have different sensitivities to broken rotor and bearing faults, a multi-channel deep convolutional neural network (MC-DCNN) fault diagnosis model based on multi-source signals is proposed in this paper, which integrates the original signals of vibration and current of the motor. Dynamic attenuation learning rate and SELU activation function were used to improve the network hyperparameters of MC-DCNN. The dynamic attenuated learning rate can improve the stability of model training and avoid model collapse effectively. The SELU activation function can avoid the problems of gradient disappearance and gradient explosion during model iteration due to its function configuration, thereby avoiding the model falling into local optima. Experiments showed that the proposed model can effectively solve the problem of motor compound fault identification, and three comparative experiments verified that the improved method can improve the stability of model training and the accuracy of fault identification.

Due to harsh and variable working conditions, the wind turbine gearbox may be damaged during the operation, resulting in an extended downtime with reduced productivity and economic loss. This calls for efficient fault diagnostics for the wind turbine gearboxes. Commonly-used diagnostics based on classical deep learning networks cannot guarantee good performance with time series signals due to the weakness of feature extraction. For this reason, an efficient channel attention residual learning approach is proposed to enhance the feature extraction and fault diagnosis for wind turbine gearboxes, leading to the development of a channel attention residual network (CAResNet). The collected time series signals are directly employed as the input. The efficient channel attention is embedded into the residual network to capture features and improve fault diagnosis capability. Experiments are carried out on a real wind turbine gearbox. The results showed that the average diagnostic accuracy of the CAResNet model was 94.41%. CAResNet has better diagnostic accuracy than other methods proposed in this paper.

Convolutional neural network (CNN) was widely applied to the data-driven based fault diagnosis. However, it often needs to artificially transform the signal into a two-dimensional (2D) image with the help of time-frequency transformation, further, the alternative one dimensional (1D) convolutional neural network can only extract single scale features but cannot adaptively reveal the relationship between scales even if 1D convolution kernels of multiple scales are applied at the same time. Accordingly, this paper proposes a mixed convolutional neural network model, namely a multiscale dynamic weighted 1D to 2D (1D-2D) convolutional neural network model (1D-2D MDWCNN), which resembles multi-wavelet based CNN method but uses different scale convolution kernels in 1D convolution neural network to facilitate the multiscale feature extraction, and accomplish adaptive features fusion in the constructed 1D-2D seamless joint network framework. Specially, to reflect the global contribution of different convolution kernels, a dynamic weighted network layer is constructed to adaptively adjust the global weight of each convolution kernel, so as to improve the fault diagnosis ability. The validity of the model is verified by motor bearing fault diagnosis experiment and gearbox fault diagnosis experiment. The diagnosis results proved that the developed 1D-2D MDWCNN model superior to the latest CNN based models.

An electric motor is a paramount part of rotating machinery, which plays an important role in industrial applications. The motor is an electromechanical coupling system, prone to electromagnetic mechanical coupling failure. It is difficult for traditional fault diagnosis methods based on a single signal source to comprehensively extract features of composite faults and ensure reliable results, as composite fault signals are coupled and interfere with each other. Recently, fault diagnosis methods leveraging multi-source heterogeneous information fusion have gained significant attention. Traditional information fusion techniques employ fixed-weight fusion strategies but suffer from a lack of dynamic adaptability. Although attention mechanism-based fusion enables dynamic feature weighting to overcome this limitation, its opacity and lack of physical interpretability remain challenges. To address these challenges, this paper proposes a novel model for electric motor compound fault diagnosis, termed the multi-source heterogeneous data adaptive feature fusion deep convolutional neural network. This model incorporates a multi-channel architecture that effectively preventing the mixing of feature information. The feature weights of different signals during data fusion are dynamically adjusted through an interpretable adaptive fusion strategy to capture more comprehensive fault features. The dropout method is employed to reduce redundancy and enhance the extraction of quality features from different signal sources. Finally, experimental results on fault detection on the experimental platform show that the accuracy of this method has reached 99.42%, demonstrating superior fault diagnosis capability, proving the feasibility and effectiveness of the method.

Discriminative feature extraction is a challenge for data-driven fault diagnosis. Although deep learning algorithms can automatically learn a good set of features without manual intervention, the lack of domain knowledge greatly limits the performance improvement, especially for nonstationary and nonlinear signals. This paper develops a multiscale information fusion-based stacked sparse autoencoder fault diagnosis method. The autoencoder takes advantage of the multiscale normalized frequency spectrum information obtained by dual-tree complex wavelet transform as input. Accordingly, the multiscale normalized features guarantee the translational invariance for signal characteristics, and the stacked sparse autoencoder benefits the unsupervised feature learning and ensures accurate and stable diagnosis performance. The developed method is performed on motor bearing vibration signals and worm gearbox vibration signals, respectively. The results confirm that the developed method can accommodate changing working conditions, be free of manual feature extraction, and perform better than the existing intelligent diagnosis methods.