传动系
相(物质)
估计
控制理论(社会学)
汽车工程
扭矩
计算机科学
工程类
物理
人工智能
量子力学
热力学
系统工程
控制(管理)
作者
Bernard Steyaert,Ananda Tjakra Adisurja,Matthias Preindl
标识
DOI:10.1109/apec48143.2025.10977308
摘要
A methodology for computing the physical parameters of a two-level VSI three-phase PMSM drivetrain is described. The method uses a load test at constant speed, and lumps resistive voltage drop, core loss resistance, and inverter nonlinearity together. Motor inductance, flux offset, and drivetrain resistance are calculated for a group of setpoints. The method uses a constant speed load test with only current sensors and a position sensor, and overcomes the rank deficiency problem by including adjacent points in an overdetermined unconstrained least squares minimization problem. Results demonstrate satisfactory physical parameter estimation with a 15% increase in airgap torque estimation accuracy compared to nameplate while withstanding a robustness to ±20% speed errors and ±10% current space harmonics.
科研通智能强力驱动
Strongly Powered by AbleSci AI