Machine Learning Based Fault Diagnosis for Stuck-at Faults and Bridging Faults

This paper presents a fault diagnosis method using a machine learning technique. The method neither needs to perform fault simulation nor it needs to store fault dictionaries in deducing candidate faults. The output responses of a circuit under diagnosis are applied to a trained neural network, and...

Full description

Saved in:
Bibliographic Details
Published in2022 37th International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) pp. 477 - 480
Main Authors Higami, Yoshinobu, Yamauchi, Takaya, Inamoto, Tsutomu, Wang, Senling, Takahashi, Hiroshi, Saluja, Kewal K.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.01.2022
Subjects
Bridging faults
Computers
Dictionaries
Fault diagnosis
Machine learning
Neural networks
Stuck-at faults
Training
Training data
Online AccessGet full text

Cover

More Information
Summary:This paper presents a fault diagnosis method using a machine learning technique. The method neither needs to perform fault simulation nor it needs to store fault dictionaries in deducing candidate faults. The output responses of a circuit under diagnosis are applied to a trained neural network, and candidate faults are obtained as a result. The paper also investigates the generation of data that are used to train the neural network. The effectiveness of the proposed method is shown by the experimental results for benchmark circuits.
DOI:10.1109/ITC-CSCC55581.2022.9894966