Senzor pro analýzu rozptylového pole asynchronního motoru

Abstract

This thesis deals with the design, implementation, and experimental validation of a sensor for the non-invasive diagnosis of an induction motor, with a focus on detecting inter-turn short circuits in the stator windings. The diagnosis is based on spectral analysis of the axial component of the stray magnetic field, which arises when the electromagnetic symmetry of the machine is disrupted. The theoretical section describes the fundamentals of asynchronous motors, the physical principles of a stray magnetic field, an overview of diagnostic methods, and the specifics of coil sensor design. In the practical section, a sensor is designed and implemented in the form of an air-core inductor integrated into a 3D-printed structure, and its properties have been experimentally verified in terms of electrical and frequency parameters. Measurements were performed on an asynchronous motor with a simulated inter-turn short circuit and evaluated using FFT algorithm in the MATLAB environment. The results show significant changes in the frequency spectrum, particularly an increase in the fundamental harmonic component and the emergence of even harmonics. The ratio of the amplitudes of the fundamental and third harmonic components was determined to be the most robust diagnostic indicator, exhibiting high sensitivity to faults while maintaining good resistance to changes in operating conditions. The proposed sensor represents a structurally simple and effective solution for the non-invasive diagnosis of induction motors.