Stator Inter Turn Fault Detection in Inverter Fed Induction Motor Using Novel DeepCGS Model

View at publisher Summarise with VIRAT
Authors: Fatimatelbatoul Mahmoud Husari, Jeevanand Seshadrinath
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Monitoring and diagnostics of inverter-fed induction motors is important since it is an inevitable part of smart grid, renewable energy systems and other critical applications. After the advent of deep learning-based techniques, in which the feature descriptors are extracted directly from the raw signals, these methods are finding place in such applications, in recent times. Thus, the merits of machine learning and deep learning techniques have been exploited in this study for early diagnosis of inter turn short circuit fault in inverter-fed induction machine, in presence of different speed and load conditions. A novel deep trihybrid DeepCGS structure combining 1D convolutional neural network (1DCNN), gated recurrent unit (GRU) and support vector machine (SVM) has been proposed for early fault detection in drive-fed induction motor. The useful spatial and temporal features are extracted by the trained 1DCNN-GRU model in the first stage. Then, these features are fed to SVM for classification. The proposed scheme has been compared to state-of-art techniques, and the experimental evaluation reveals a better performance as per the performance parameters such as accuracy, specificity and sensitivity. To prove the effectiveness of the proposed trihybrid DeepCGS architecture, the comparable performances of the classifiers have been verified using t-test and Friedman test. The results establish that the proposed trihybrid model “deepCGS” is statistically significant, and the results are not due to mere chance.

Fields of Study & Subjects

Machine Fault Diagnosis Techniques Sensorless Control of Electric Motors Electric Motor Design and Analysis

Publication details

Journal / source
IEEE Transactions on Industry Applications
Pages
1–9
Publication date
1 January 2026
Language
EN