Answers are grounded in this paper and cite their sources.
Investigates fault detection in induction machines, suggesting effective classification methods in challenging environments.
Industrial induction machines are designed to achieve high performances under harsh conditions. They often operate near magnetic saturation. In this operating condition, the early detection of electrical and mechanical faults is a major challenge for the safety operation of the machine. This study proposes a complete pipeline of faults detection and classification of a saturated tree phase induction machine based on Finite Element model and Machine learning algorithms with model selector. Eleven faults are simulated under saturated regime and their influences on the machine parameters such as the electromagnetic torque; the voltage; the current; and the rotor speed are analyzed. During faults simulation, four frequency features (dominant Frequency, Spectral Energy, Spectral Centroid, Total Harmonic Distortion) are extracted and recorded in a dataset. Multiple machine learning algorithms are trained with rigorous statistical evaluation protocol on the dataset to correctly classified each type of fault. The study end with a noise analysis performed to test the robustness of the model. The obtained results show that, Tree-based ensemble methods show highest performance in a clean and very noisy environment with accuracies ranging from 70% for a very noisy environment to 99% for clean environment.
No takes yet. Share an insight, caveat, or question.
Yapele et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: