Small sample fault diagnosis for wind turbine gearbox based on lightweight multiscale convolutional neural network

Abstract

The maintenance and diagnosis of wind turbine gearboxes are crucial for enhancing the stability and operational efficiency of wind power systems. However, there are still two challenges in gearbox fault diagnosis methods based on deep learning : (1) Limited failure sample; (2) Interference of strong noise. To solve the above issues, a lightweight multiscale convolutional neural network (LMSCNN) based fault diagnosis method is proposed in this paper. Among them, a large kernel convolution is used to denoise the original vibration signal. A lightweight multiscale architecture is constructed using depthwise separable convolutional blocks (DSCBs), which mine fault features at different scales and improve the operational efficiency of the model. Moreover, a parallel global pooling block (PGPB) is designed to provide a more comprehensive feature for the fusion layer, enabling the effective diagnosis of vibration signals. Experiments are conducted on the datasets of two different gearboxes, which prove that LMSCNN has excellent generalization capability and diagnostic speed.

A novel multi-sensor hybrid fusion framework

“…Researchers usually adopt strategies to address these issues, such as adding multiple convolutional channels of various scales to the model. This approach aims to enhance the model's ability to perceive global information and improve robustness [48,49]. In addition, the attention mechanism can be introduced to enhance the CNN's ability to extract key features [50,51].…”

Data-Driven Models Applied to Predictive and Prescriptive Maintenance of Wind Turbine: A Systematic Review of Approaches Based on Failure Detection, Diagnosis, and Prognosis

“…Ref. [81] proposed an intelligent fault diagnosis method for wind turbine gearboxes using a multiscale convolutional neural network-LMSCNN (Figure 13). The model was able to learn resources with the direct use of the vibration signal without any type of pre-processing, with a high capacity for versatility and excellent prospects for applications in industrial centers.…”

Intelligent fault diagnosis for electro-hydrostatic actuator based on multisource information convolutional residual network

“…This innovative approach is capable of automatically learning high-level nonlinear features from raw sample inputs, effectively bypassing the constraints posed by traditional shallow machine-learning algorithms that are heavily reliant on feature engineering [16]. In particular, the convolutional neural network (CNN), one of the most popular DL methods, has become a significant focus in the realm of fault diagnosis research for a variety of rotating machinery, including bearings [17,18], gearboxes [19], and hydraulic pumps [20,21].…”