IInception-CBAM-IBiGRU based fault diagnosis method for asynchronous motors

Area of Science:

• Electrical Engineering
• Artificial Intelligence
• Machine Learning

Background:

• Traditional deep learning methods struggle with extracting robust asynchronous motor fault features, especially in noisy environments.
• Accurate fault diagnosis is crucial for maintaining the operational integrity and longevity of asynchronous motors.

Purpose of the Study:

• To propose an end-to-end fault diagnosis method for asynchronous motors that overcomes limitations of existing deep learning approaches.
• To enhance the accuracy and robustness of asynchronous motor fault diagnosis, particularly under adverse noise conditions.

Main Methods:

• A novel method combining signal-to-grayscale image conversion, Improved Inception (IInception) modules, Convolutional Block Attention Module (CBAM), and Improved Bi-directional Gate Recurrent Unit (IBiGRU).
• Feature extraction using 2D convolution, multi-scale learning with IInception residual blocks, attention-based feature refinement with CBAM, and temporal feature extraction with IBiGRU.
• Hyperparameter optimization using the Weighted Mean Of Vectors Algorithm (INFO) and fault identification via the SoftMax function.

Main Results:

• The proposed IInception-CBAM-IBiGRU method achieved a fault diagnosis accuracy close to 100% for asynchronous motors.
• The method demonstrated significant robustness and maintained high accuracy even in low signal-to-noise ratio environments.
• Experimental results validated the effectiveness and generalization ability of the proposed fault diagnosis approach.

Conclusions:

• The developed end-to-end fault diagnosis method offers a highly accurate and robust solution for asynchronous motor fault detection.
• The integration of IInception, CBAM, and IBiGRU effectively addresses the challenges of feature extraction and diagnosis in noisy conditions.
• This approach significantly improves the reliability and performance of asynchronous motor monitoring systems.