An End-to-End Fault Diagnosis Model for Rolling Bearings Based on Multi-Scale Convolution and the Kolmogorov–Arnold Network

Donghua Yu, Zhenyu Wang, Jia Liu, Huan Liu, Changtian Ying · 2026

Read An End-to-End Fault Diagnosis Model for Rolling Bearings Based on Multi-Scale Convolution and the Kolmogorov–Arnold Network by Donghua Yu, Zhenyu Wang, Jia Liu, Huan Liu, Changtian Ying (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13364184.txt.

Description

Donghua Yu and colleagues present a 2026 study on spotting rolling bearing faults straight from raw vibration signals, with no hand-built features. Their KanMSConv model stacks multi-scale depthwise convolutions that catch both sharp fault pops and slow modulations. It adds channel attention to stress the telling parts. It also swaps the usual classifier head for a Kolmogorov-Arnold network that bends flexibly around complex fault borders.

On three public bearing datasets spanning loads and machines, the model scored at or near perfect accuracy, topping mainstream rivals on every score reported. Ablation runs confirmed each module adds value, and cost checks kept the design practical. The authors pitch the end-to-end route for factory lines that cannot afford manual signal work.

Catalog record

CREATORDonghua Yu, Zhenyu Wang, Jia Liu, Huan Liu, Changtian Ying
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE55652
LICENSEhttps://creativecommons.org/licenses/by/4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/article/MED/PMC13364184

Browse the full library · About the Tower of Babel