Search

When using multi-sensor information fusion technology to evaluate the health status of motors, there are outliers and missing values in the monitoring data of mine motors. However, deep learning models such as convolutional neural networks and recurrent neural networks find it difficult to effectively extract data features and update network weights when the data quality is severely degraded, resulting in problems such as vanishing or exploding gradients. In order to solve the above problems, A CNN-GRU (CNN-GRU-AMSA) model based on adaptive multi-scale attention mechanism is proposed to evaluate the health status of mine motors. Firstly, the model fills in, removes, and standardizes the motor operation data collected by sensors, and classifies the operating conditions of mine motors based on environmental temperature changes. Secondly, based on the Mahalanobis distance, the health index (HI) of health evaluation indicators such as motor current, three-phase temperature of motor winding, front bearing temperature of motor, and rear bearing temperature of motor are calculated. The Savitzky Golay filter is used to denoise, smooth, and normalize the HI indicator. Combining the contribution of different indicators calculated by principal component analysis method to mine motors, the weighted fusion of indicator HI is used to obtain the mine motor HI. Finally, the mine motor HI is input into the CNN-GRU-AMSA model, which dynamically adjusts attention weights to achieve information fusion of features at different scales, thereby accurately outputting the health status evaluation results of the motor. The experimental results show that compared with other common deep learning models such as CNN, CNN-GRU, CNN-LSTM, and CNN-LSTM Attention, the CNN-GRU-AMSA model performs better in evaluation metrics such as root mean square error, mean absolute error, accuracy, Macro F1, and MicroF1. The model has a smaller fluctuation range and better stability in predicting residuals.

In order to detect abnormal wind turbine generator and reduce the occurrence of outages, a deep learning framework combining K-CNN and N-GRU is proposed based on multi-dimensional sensor parameters recorded in real wind farm SCADA system, and a wind turbine generator state prediction model is established. Firstly, the correlation of state parameters was analyzed by Pearson correlation coefficient, and then the one-dimensional fusion parameters were weighted by weight coefficient. Secondly, to solve the problem of ignoring shallow features in traditional feature extraction, CNN was used to extract the features of one-dimensional fusion parameters in layers, and KPCA was used to reduce the feature extraction results of different layers to one dimension. Then, to solve the problem of parameter optimization of the traditional GRU algorithm, the neural network architecture search was used to improve the GRU algorithm, and the N-GRU model was obtained. The feature extraction results after dimensionality reduction were input into N-GRU for prediction and reconstruction error was obtained, then the state evaluation was realized by setting the alarm threshold. Finally, a 2 MW wind turbine in a wind farm in Xinjiang was taken as an example to verify the model validity and model accuracy.

Objective To investigate the role of ski-related novel protein N (SnoN) in the pathogenesis of intrauterine adhesion (IUA) and its association with transforming growth factor-β1 (TGF-β1) and Smad4 in endometrial tissue in IUA patients. Methods Endometrial tissues of 20 severe IUA patients (sIUA group) and 20 healthy women with uterine septum (control group) were collected from our 2 departments during January 2019 and October 2020. Immunohistochemical assay, Western blotting and reverse transcription quantitative polymerase chain reaction (RT-qPCR) were used to detect the expression of SnoN, TGF-β1 and Smad4 in the endometrial tissue of the control and sIUA groups. The correlations of their expression levels in the endometrial tissues were analyzed with Spearman rank correlation analysis to investigate the association of the levels and IUA. Results The expression levels of TGF-β1 and Smad4 were significantly higher, while that of SnoN was obviously lower in the sIUA group than the control group (all P < 0.01). In the endometrial tissues from the sIUA group, the expression of Smad4 was positively correlated with that of TGF-β1 (r=0.468 9, P < 0.05). However, the TGF-β1 expression was negatively correlated with that of SnoN (r=-0.483 5, P < 0.05). Conclusion SnoN may be involved in the pathogenesis of IUA via the TGF-β1/Smad4 signal pathway.

To more precisely assess the reliability of wind turbine gears, a competing failure model is established with consideration given to the degradation of internal fatigue strength and external random shocks. Firstly, the nonlinear Gamma process is adopted to depict the degradation of fatigue strength, and the strength‐stress interference theory is used to establish the threshold of degradation failure that changes over time. Also, the non‐homogeneous Poisson process is employed to depict the arrival of the shock, and classification of the shocks. Secondly, the correlation between degradation and shock is described, while the effect of shocks on degradation is demonstrated as a sudden increase in the extent of degradation. Besides, the decline rate of residual strength curve is introduced to describe the change in intensity parameters of the Poisson process, and then the effect of degradation on shocks is described. Finally, the sun gear in the wind turbine gearbox is an example. Monte Carlo simulation is used to calculate the reliability of the dependent competing failure model. This is then compared with the conventional stress‐strength interference model. The results verify that when the relationship between degradation and the shock process is disregarded, the reliability of gears may be overestimated. [ABSTRACT FROM AUTHOR]

A novel method is proposed to investigate the pattern of variation in the residual strength and reliability of wind turbine gear. First, the interaction between loads and the effect of the loading sequence is considered based on the fatigue damage accumulation theory, and a residual strength degradation model with few parameters is established. Experimental data from two materials are used to verify the predictive performance of the proposed model. Secondly, the modeling and simulation of the wind turbine gear is conducted to analyze the types of fatigue failures and obtain their fatigue life curves. Due to the randomness of the load on the gear, the rain flow counting method and the Goodman method are employed. Thirdly, considering the seasonal variation of load, the decreasing trend of gear fatigue strength under multistage random load is calculated. Finally, the dynamic failure rate and reliability of gear fatigue failure under multistage random loads are analyzed. The results demonstrate that the randomness of residual strength increases with increasing service time. The seasonality of load causes fluctuations in the reliability of gear, providing a new idea for evaluating the reliability of the wind turbine gear. [ABSTRACT FROM AUTHOR]

Accurate status prediction is critical for detecting potential failures of Doubly‐fed induction generator (DFIG). A unique Digital Twin (DT) fault prediction paradigm is put forth in this research, which uses an edge intelligence paradigm to ensure high‐performance fault prediction. Edge‐based DT provides compute and storage capabilities on edge devices to enable effective data processing. The stator current spectrum and instantaneous power spectrum of the DFIG are simulated and assessed, followed by the development of a verification prototype on the DFIG and the qualitative observation of the internal performance of the DFIG. The encoder is constructed using Squeeze‐and‐Excitation (SE) Convolutional Neural Network (SE‐CNN) with the channel attention mechanism, and the decoder is constructed based on a long short‐term memory (LSTM) network to form the self‐encoding model named SE‐CNN‐LSTM. In order to forecast problems and provide a foundation for predictive maintenance, the status indicator is built based on the prediction residuals of the status variables. Based on the outcomes of edge‐cloud co‐simulation and data collected from the Supervisory Control and Data Acquisition system at the Da Bancheng Wind Farm in Xinjiang, China, it is demonstrated that the framework and technology are workable and capable of predicting generator failures. [ABSTRACT FROM AUTHOR]