Veronica Hernandez-Ramirez

Although the Space Phasor Model (SPM) is widely used to analyze induction motors, this study presents a new framework that applies SPM with geometric parameter extraction via a metaheuristic optimization algorithm. Differential Evolution was chosen for its simplicity, reliability, and effectiveness in exploring the parameter space. The Space Vector is constructed from the three-phase current signals. The approach captures spatial trajectories: healthy operation exhibits circular patterns, while inter-turn short-circuit (ITSC) faults deform them into ellipses, revealing both the affected phase and fault severity. The ellipse parameters are normalized using Z-score and organized into a structured fault database that includes both no-load and full-load conditions, supporting load-aware classification of 25 operating states and substantially extending the 13-class scope of previous studies. Tests with multiple classifiers (including Random Forests, Extra Trees, and gradient boosting) confirm the approach’s reliability, achieving 98.7% accuracy across 25 classes, 99.5% for the 13 most representative faults, and 94.1% under full-load for seven classes. Overall, this approach offers a phase-specific, severity-oriented diagnostic tool, highlighting its potential for industrial motor fault analysis and setting a new benchmark for multi-class classification.

Automatic detection of early-stage electrical faults in motors is challenging because physical signals, such as temperature or vibrations, are often linked to regular machine operation. Inter-turn short-circuit (ITSC) faults in the stator winding are a leading cause of irreversible damage. Given industrial demands, there is an ongoing need for innovative approaches that reduce dimensionality and expedite feature extraction from three-phase current signals. This work presents an ITSC multi-fault classification algorithm by modeling the behavior of three-phase stator currents in a complex space. To achieve this, two geometric-based and two optimization-based approaches are employed to parameterize the shape of the complex signal. Rather than directly extracting features from the raw signals, these parameters capture the similarities among incipient failures. The algorithm is evaluated using machine learning classifiers trained on a dataset generated from an experimental test bench and a publicly available dataset. The proposed method achieved an accuracy of 95.30% across 13 categories, demonstrating its robustness and reliability and positioning it as a highly competitive alternative to state-of-the-art techniques.