Multiphysics Modeling and Optimal Current Profiling for Switched Reluctance Machine Drive

Bingchu Li,Xiao Ling,Yixiang Huang,Liang Gong,Chengliang Liu
DOI: https://doi.org/10.1109/iecon.2018.8591754
2018-01-01
Abstract:The aim of this paper is to realize multi-objective optimal control of a switched reluctance machine considering three criteria, namely, torque ripple, vibration (radial force)and RMS current. A harmonic model is used during optimization, which represents current profile, radial force factor and torque factor in Fourier series. Multi-physics model is built based on harmonic model, in which the three criteria are linear functions of harmonic amplitude contained in phase currents. A collaborative optimization is proposed to obtain current profile considering all three criteria. Weight coefficients are assigned to three criteria to guide trade-off during optimization. Current profiling is implemented under four modes, which put emphasis on different criteria. The torque ripple, radial force and RMS current under four modes are compared through simulation and experiments. It is demonstrated that the three criteria are effectively adjusted through regulating weight coefficients, the proposed method gives optimal current profiles with designated compromise between multiple performance criteria and avoids some criteria being severely deteriorated while others are optimized.
What problem does this paper attempt to address?