PHM for Rotating Machines

Published in: Measurement (JIF: 5.2, Rank: 9.1%)

Publication Date: November 2023

Paper Link : https://www.sciencedirect.com/science/article/pii/S0263224123010102

💡 Simple Summary

A PMSM(Permanent Magnet Synchronous Motor) fault diagnosis method is proposed that effectively handles variable operating conditions, mitigating frequency and amplitude changes in stator current signal while emphasizing small fault-related residuals.

Abstract

This paper proposes a novel preprocessing method, namely stator current operation compensation (SCOC), for deep learning-based fault diagnosis of a permanent magnet synchronous motor (PMSM) under variable operating conditions. To solve the problem that a change in operating conditions modifies characteristics of stator current signals in a way that makes it hard to discriminate fault modes, SCOC includes two stages: 1) tacho-less resampling and 2) main operating component subtraction with rescaling. First, the stator current signal is resampled to contain the same cycle for each deep learning input. Next, the signal is transformed to alleviate amplitude change by torque and to emphasize relatively small fault components. The effectiveness of the proposed method was validated by applying it to experimental data acquired under different types of PMSM operations. The results showed that each SCOC stage can reduce the variability between the data and effectively increase the fault diagnosis performance of PMSMs.

Published in: Journal of Computational Design and Engineering (JIF: 4.8, Rank: 11.3%)

Publication Date: October 2023

Paper Link : https://www.sciencedirect.com/science/article/pii/S0263224123010102

💡 Simple Summary

A CNN-based fault diagnosis framework is proposed, enhancing feature learning through multi-scale structures with assigning adaptive weights to different temporal scales 

Abstract

Multi-scale convolutional neural network structures consisting of parallel convolution paths with different kernel sizes have been developed to extract features from multiple temporal scales and applied for fault diagnosis of rotating machines. However, when the extracted features are used to the same extent regardless of the temporal scale inside the network, good diagnostic performance may not be guaranteed due to the influence of the features of certain temporal scale less related to faults. Considering this issue, this paper presents a novel architecture called a multi-scale path attention residual network to further enhance the feature representational ability of a multi-scale structure. Multi-scale path attention residual network adopts a path attention module after a multi-scale dilated convolution layer, assigning different weights to features from different convolution paths. In addition, the network is composed of a stacked multi-scale attention residual block structure to continuously extract meaningful multi-scale characteristics and relationships between scales. The effectiveness of the proposed method is verified by examining its application to a helical gearbox vibration dataset and a permanent magnet synchronous motor current dataset. The results show that the proposed multi-scale path attention residual network can improve the feature learning ability of the multi-scale structure and achieve better fault diagnosis performance.

Published in: (Paper Under Review)

Publication Date: (Paper Under Review)

Paper Link : (Paper Under Review)

💡 Simple Summary

An uncertainty-aware Fault Diagnosis method is proposed using a specific classifier structure with loss function which is effective in detecting unseen faults.

Abstract

(confidential)

Published in: (Paper in progress)

Publication Date: (Paper in progress)

Paper Link : (Paper in progress)

💡 Simple Summary

A motor severity estimation method that demonstrates superior performance in previously unseen severity regions by leveraging simulation data has been developed. 

Abstract

(confidential)