Evaluating the Performance of Generative Adversarial Network, Synthetic Minority Oversampling Technique, and Adaptive Synthetic Sampling in Marine Diesel Engine Fault Diagnosis using Vibration Data

Principal component analysis dimensionality reduction analysis demonstrates that the samples generated by the three data augmentation algorithms are uniformly distributed around the real samples, with distinct boundaries, thereby preserving the authenticity of the generated data.

Authors

Zhongjun Wang
Xiong Bao
Xiaofeng Guo (ORCID: https://orcid.org/0000-0003-1971-7442)
Yingli Qu (ORCID: https://orcid.org/0000-0001-6685-8067)
Jie Wang (ORCID: https://orcid.org/0000-0002-5198-9563)
Pengcheng Wang (ORCID: https://orcid.org/0000-0001-6395-4870)
Zhijun Chen (ORCID: https://orcid.org/0000-0003-4761-6710)
Ziyu Xiao

Publication Details

Journal: Engineered Science
Published: 2025-01-01
DOI: https://doi.org/10.30919/es1696
Primary Topic: Machine Fault Diagnosis Techniques
Type: article
Field-Weighted Citation Impact: 0.00

Evaluating the Performance of Generative Adversarial Network, Synthetic Minority Oversampling Technique, and Adaptive Synthetic Sampling in Marine Diesel Engine Fault Diagnosis using Vibration Data

Authors

Publication Details

Evaluating the Performance of Generative Adversarial Network, Synthetic Minority Oversampling Technique, and Adaptive Synthetic Sampling in Marine Diesel Engine Fault Diagnosis using Vibration Data

Evaluating the Performance of Generative Adversarial Network, Synthetic Minority Oversampling Technique, and Adaptive Synthetic Sampling in Marine Diesel Engine Fault Diagnosis using Vibration Data

Abstract

Ask Laika to Summarize, Analyze, and Connect papers live on the map.