Analysis and Detection of Rotor Eccentricity in Permanent Magnet Synchronous Machines Based on Linear Hall Sensors

Rotor eccentricity commonly exists in permanent magnet synchronous machines (PMSMs) due to improper assembly, excessive vibration, and continuous operation. Consequently, the unbalanced magnetic pull, introduced by rotor eccentricity, may lead to unfavorable noise, vibration, and even bear damage, which makes the detection of eccentricity meaningful. Although motor current signal analysis is the most popular method to diagnose eccentricity based on the amplitudes of sideband component, it is a big challenge to classify different types of eccentricities and quantify the eccentric ratio due to the spatiotemporal coupling effect. In this article, by installing a set of Hall sensors at stator slots of PMSMs to measure the magnetic flux density in time domain, the static eccentricity and dynamic eccentricity can be easily separated. This article provides a novel complex-analytical model to reveal the single side band modulation effect in air-gap filed due to dynamic eccentricity. Moreover, a diagnosis algorithm based on frequency-adaptive notch filter is designed to identify the eccentricity type and the corresponding degree. The effectiveness of the novel diagnosis method is validated by both simulations and experiments.

Automated summary

This article introduces a method to separate static eccentricity and dynamic eccentricity by installing Hall sensors to measure magnetic flux density. It also proposes a diagnosis algorithm based on frequency-adaptive notch filter to identify the type and degree of eccentricity.