Novel Machine Parameter Estimation Scheme Toward Accurate Maximum Torque Production for Dual Three-Phase PMSMs

Ze Li,Guodong Feng,Wenlong Li,Narayan C. Kar,Donovan O'Donnell

DOI: https://doi.org/10.1109/tte.2021.3083667

IF: 6.519

2021-12-01

IEEE Transactions on Transportation Electrification

Abstract:This article proposes a hybrid signal injection-based maximum torque per ampere (MTPA) control for dual three-phase permanent magnet synchronous machines (PMSMs). The proposed method consists of a real dc current injection in the harmonic subspace and a virtual ac current injection in the fundamental subspace. The real dc current injection in the harmonic subspace does not impact the average torque production, and induced losses are small enough to be neglected. In the meantime, to guarantee the accuracy of the proposed MTPA method, a novel machine parameter estimation scheme considering voltage source inverter nonlinearity effects is proposed in this article, which can track the variation of $q_{1}$ -axis inductance and resistance accurately. Therefore, the proposed method only requires command voltages and currents in the torque model, and the variations of machine parameters are compensated by the proposed parameter estimation scheme. As the command voltages have correspondence with the variations of machine parameters, the magnetic saturation and temperature effects have been taken into account. The proposed method is validated by simulations and experimental results under various torque, speed, and temperature conditions. Additionally, the accuracy of the proposed method is validated, and its performance is proven to be superior through comparison of the derived results with those obtained from various existing methods.

engineering, electrical & electronic,transportation science & technology