Analysis on accuracy of rotor‐locked transient‐response‐based vibration model for switched reluctance machines

• Published in: IET electric power applications Vol. 16; no. 5; pp. 616 - 633
• Main Authors: Li, Bingchu, Dong, Zikang, Huang, Yixiang, Liu, Chengliang
• Format: Journal Article
• Language: English
• Published: Wiley 01.05.2022
• Subjects: magnetic forces; reluctance machines; vibrations
• ISSN: 1751-8660; 1751-8679
• DOI: 10.1049/elp2.12181

This paper investigates the accuracy of rotor‐locked transient‐response (RLTR)‐based vibration model for switched reluctance machines (SRM). RLTR is stator vibration stimulated by abrupt change of radial force when turning off one phase of SRM with rotor being locked, the total vibration of SRM can be modelled as linear superposition of RLTR series. An electromechanical model is built to clarify the generation of SRM vibration, it is revealed that some velocity‐depend terms in electromechanical model would be neglected when rotor is locked, it leads to inevitable difference of RLRT between rotor‐locked condition and rotational condition, the difference is illustrated through multi‐physics simulation. A harmonic radial force model is proposed to quantitatively evaluate the accuracy of RLTR‐based vibration model, radial force spectrums with locked rotor and rotational rotor are compared to predict relative error between RLTR‐based vibration and actual vibration. The influence of turn‐off angle and velocity on accuracy of RLTR‐based vibration model is also analysed. The accuracy of RLTR‐based vibration model is evaluated through experiment on 3‐phase 12/8 SRM, the results showed that: (1) the accuracy of RLTR‐based vibration model is impacted by turn‐off angle and velocity; (2) relative error between RLTR‐based vibration and actual vibration can be predicted using harmonic radial force model. This paper illustrates the generation mechanism of modelling error for commonly used RLTR‐based vibration model of SRM, the proposed error prediction method could lead to improved accuracy on SRM vibration model through error compensation.