This paper proposes a novel intelligent fault diagnosis method to automatically identify different health conditions of wind turbine gearbox. Different from traditional approaches where feature extraction and classification are separately designed and performed, this study aims to automatically learns effective fault features directly from raw vibration signals and meanwhile classifies the type of faults in a single learning framework, thus leading to an end-to-end learning based fault diagnosis system for wind turbine gearbox without additional signal processing techniques and diagnostic expertise. Considering multi-scale characteristics inherent in vibration signals of a gearbox, in this paper, a new multiscale convolutional neural network (MSCNN) framework is proposed to perform multi-scale feature extraction and classification simultaneously. A key advantage of the proposed MSCNN is the ability to learn complementary and rich fault pattern features at different time scales by incorporating multiscale learning into the CNN architecture, which greatly improves feature learning to enable better diagnosis performance. The proposed MSCNN framework is evaluated through experiments on a wind turbine test rig. The results demonstrate that the proposed MSCNN framework outperforms state-of-the-art methods and exhibits superior robustness against noise.

The supervisory control and data acquisition (SCADA) systems are widely installed on wind turbines (WTs) in major wind farms, which produce a large amount of sensory data that can be used for fault diagnosis of WTs. However, these SCADA data are naturally multivariate time series, which represent complex temporal correlations within each sensor variable and spatial correlations between different sensor variables. To effectively capture spatio-temporal correlations in SCADA data, we propose a new spatio-temporal multiscale neural network (STMNN). The proposed STMNN model contains two parallel feature extraction modules: first, a multiscale deep echo state network module to extract temporal multiscale features; second, a multiscale residual network module to extract spatial multiscale features. Additionally, as the WTs are in a normal working state most of the time, there are a large amount of normal SCADA data and few failure SCADA data. To address the data imbalance problem of SCADA data and enhance the fault diagnosis performance, instead of cross-entropy loss, the STMNN model adopts focal loss as loss function. Our proposed STMNN method can provide an end-to-end fault diagnosis solution with imbalanced SCADA data, and is evaluated through experiments on an SCADA dataset from a real wind farm. The experimental results and comparative analysis have proved the effectiveness of our proposed STMNN model in practical applications.

Brain-computer interface (BCI) based on motor imagery (MI) can control external applications by decoding different brain physiological signals, such as electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS). Traditional unimodal-based MI decoding methods cannot obtain satisfactory classification performance due to the limited representation ability in EEG or fNIRS signals. Usually, different brain signals have complementarity with different sensitivity to different MI patterns. To improve the recognition rate and generalization ability of MI, we propose a novel end-to-end multi-modal multi-task neural network (M2NN) model with the fusion of EEG and fNIRS signals. M2NN method integrates spatial-temporal feature extraction module, multimodal feature fusion module, and multi-task learning (MTL) module. Specifically, the MTL module includes two learning tasks, namely, one main classification task for motor imagery and one auxiliary task with deep metric learning. This approach was evaluated using a public multi-modal dataset and experimental results show that M2NN achieved the classification accuracy improvement of 8.92%, 6.97%, and 8.62% higher than unimodal multi-task MEEG, MHbR, and multi-modal single-task MDNN, respectively. And classification accuracies of multi-task methods of MEEG, MHbR, and M2NN are improved by 4.8%,4.37%, and 8.62% compared with single-task methods EEG, HbR, and MDNN. M2NN method achieved the best classification performance of six methods, with the average accuracy of 29 subjects being 82.11% ± 7.25%. The effectiveness of multimodal fusion and multitask learning was verified. And M2NN method is superior to baseline and SOTA methods.

Effective fault detection of wind turbines (WTs) can greatly help improve their availability and reduce their operation and maintenance costs. In this context, data-driven fault detection approaches have attracted a lot of interests due to the availability of a large amount of monitoring sensor data containing rich information related to health conditions of WTs. However, sensor data collected from WTs are naturally multivariate and highly nonlinear correlated with redundant information and significantly contaminated measurement noise, which makes the WT fault detection more challenging. To this end, this paper develops a multivariate data-driven fault detection (MDFD) framework based on a recently emerged neural network algorithm named denoising autoencoder (DAE). Instead of using a single fixed noise level in the traditional DAE, a novel multi-level-denoising autoencoder (MLD-AE) method is proposed to enhance the representation learning ability by designing different multi-level noise adding schemes. The proposed MLD-AE could better discover useful patterns at multiple corrupted scales and capture nonlinear dependencies from noisy multivariate sensor data, therefore robustly reconstruct the original signal with the preserved largest information. The proposed framework and method are evaluated on both simulated data from a generic 5 MW WT benchmark and SCADA data from a real wind farm. Results demonstrate that our proposed MLD-AE-based fault detection approach significantly outperforms traditional DAE, AE and linear PCA approaches, which has great potentials for practical applications in the wind industry.

Motor dysfunction is a major consequence after stroke and it is generally believed that the loss of motor ability is caused by the impairments in neural network that controls movement. To explore the abnormal mechanisms how the brain controls shoulder abduction and elbow flexion in "flexion synergy" following stroke, we used the functional corticomuscular coupling (FCMC) between the brain and the muscles as a tool to identify the temporal evolution of corticomuscular interaction between the synkinetic and separate phases. 59-channel electroencephalogram (EEG) over brain scalp and 2-channel electromyogram (EMG) from biceps brachii (BB)/deltoid (DT) were recorded in sixteen stroke patients with motor dysfunction and eight healthy controls during a task of uplifting the arm (stage 1) and maintaining up to the chest (stage 2). As a result, compared to healthy controls, stroke patients had abnormally reduced coherence in EEG-BB combination and increased coherence in EEG-DT combination. Compared to synkinetic stroke patients, separate ones exhibited higher coupling at gamma-band during stage 1 and higher at beta-band during stage 2 in EEG-BB combination, but lower at beta-band during stage 2 in EEG-DT combination. Therefore, we infer that the disorders of efferent control and afferent proprioception in sensorimotor system for stroke patients effect on the oscillation at beta and gamma bands. Patients need integrate more information for shoulder abduction to compensate for the functional loss of elbow flexion in the recovery process, so that partial cortical cortex controlling on the elbow flexion may work on the shoulder abduction during "flexion synergy". Such researches could provide new perspective on the temporal evolution of corticomuscular interaction after stroke and add to our understanding of possible pathomechanisms how the brain abnormally controls shoulder abduction and elbow flexion in "flexion synergy".

In modern Internet of Things-enhanced wind power systems, most existing data-driven fault diagnosis approaches for wind turbines (WTs) are performed under a centralized paradigm that ignores data privacy. Recently, federated learning (FL) presented a solution to enable edge WTs located at isolated sites to collaboratively learn a shared diagnosis model without accessing local privacy-sensitive data. However, the practical issues of fault label heterogeneity among edge clients and scarcity of labeled data still severely impede the generation of a satisfactory diagnosis model. To address these issues, we propose a diagnostic knowledge-based FL framework (DKFLWT) for collaborative fault diagnosis of distributed edge WTs. In our DKFLWT framework, independently learned diagnostic knowledge from each edge client, rather than model parameters in conventional FL, is uploaded to the cloud server to enrich the client-specific information visible to the server and mitigate the adverse effects on model performance caused by label heterogeneity. To enhance the overall efficiency of the framework, we develop a two-stage, single-round training mechanism, in which the cloud server serves as a universal platform that can accommodate the customized requirements of users, implying the convenient integration of semi-supervised learning to enhance the diagnosis performance in scenarios with limited labeled data. Furthermore, a spatio-temporal memory-enhanced autoencoder is designed to sufficiently exploit essential diagnostic knowledge of different fault patterns from each client. Experimental results demonstrate superior diagnosis performance of our DKFLWT framework with an improvement of more than 22.1% in accuracy and 37.2% in training efficiency against several compared methods in all seriously heterogeneous scenarios.