Low-Computational-Cost Hybrid FEM-Analytical Induction Machine Model for the Diagnosis of Rotor Eccentricity, Based on Sparse Identification Techniques and Trigonometric Interpolation

Abstract

Since it is not efficient to physically study many machine failures, models of faulty induction machines (IMs) have attracted a rising interest. These models must be accurate enough to include fault effects and must be computed with relatively low resources to reproduce different fault scenarios. Moreover, they should run in real time to develop online condition-monitoring (CM) systems. Hybrid finite element method (FEM)-analytical models have been recently proposed for fault diagnosis purposes since they keep good accuracy, which is widely accepted, and they can run in real-time simulators. However, these models still require the full simulation of the FEM model to compute the parameters of the analytical model for each faulty scenario with its corresponding computing needs. To address these drawbacks (large computing power and memory resources requirements) this paper proposes sparse identification techniques in combination with the trigonometric interpolation polynomial for the computation of IM model parameters. The proposed model keeps accuracy similar to a FEM model at a much lower computational effort, which could contribute to the development and to the testing of condition-monitoring systems. This approach has been applied to develop an IM model under static eccentricity conditions, but this may extend to other fault types.

A new hybrid analytical model for electromagnetic analysis of wound rotor induction motors

“…2 Moreover, the user cannot obtain an in-depth knowledge of the design and performance of electric machines while using the FEM software. To this end, hybrid analytical-FEM model has been presented in [3][4][5] for real time simulation, which previously prepares 3-D look-up tables for each element of inductance matrix through magneto-static analysis by FEM. different analytical techniques have been so far introduced such as winding function theory (WFT), 6,7 magnetic equivalent circuit (MEC) model, [8][9][10] conformal mapping (CM) method, 11,12 field reconstruction method (FRM), 13,14 and sub-domain (S-D) model.…”

State-of-the-Art Techniques for Fault Diagnosis in Electrical Machines: Advancements and Future Directions

“…Model-based on FEM [68,182] • Gives accurate modeling of the electromagnetic behavior of electrical machines.…”

Desarrollo e implementación de un banco de ensayos virtual de aerogeneradores para diferentes regímenes de funcionamiento y condiciones de fallo

“…El Capítulo 5 corresponde a la publicación "Low Computational Cost Hybrid FEM-Analytical Induction Machine Model for the Diagnosis of Rotor Eccentricity, based on Sparse Identification Techniques and Trigonometric Interpolation" [131] que presenta una nueva metodología basada en la identificación dispersa y el polinomio de interpolación trigonométrica para minimizar el uso de simulaciones FEM para desarrollar un modelo híbrido FEM-Analítico de máquina de inducción con excentricidad.…”