First-Order Model Based Inductance Identification With Least Square Method for High-Speed Sensorless Control of Permanent Magnet Synchronous Machines

Yinfeng Hu,Kai Liu,Wei Hua,Mingjin Hu

DOI: https://doi.org/10.1109/tpel.2023.3263518

IF: 5.967

2023-01-01

IEEE Transactions on Power Electronics

Abstract:In sensorless control of high-speed permanent magnet synchronous machines (PMSMs), online inductance identification techniques can be employed to improve the position estimation precision. The high-order inductance identification model which is affected by the flux error, the dead-time effect equivalent voltage and initial rotor position error adopted in conventional methods often face challenges including robustness and accuracy. To address this issue, this article uses a first-order discrete-time model in the estimated -frame to identify the winding inductance with disturbance injection. The first-order model allows the identification to be independent of flux error, initial rotor position error, and inverter nonlinearity, thus the robustness is improved. The proposed method takes the computational delay and discrete-time nature of pulsewidth modulation into account to avoid the model error brought by low ratio of carrier-to-fundamental-frequency. Additionally, the results are estimated by least square method to solve fluctuation problem caused by measurement noise and the possible surge phenomenon. Experiments are carried out on a high-speed surface-mounted PMSM for vacuum cleaner to verify the proposed method.

engineering, electrical & electronic