Multi-Parameter Decoupled Estimation for PMSM Using Adaptive-Genetic-Algorithm

Zhuoxuan Yin,Guodong Feng,Yuting Lu
DOI: https://doi.org/10.1109/cac53003.2021.9727775
2021-01-01
Abstract:For permanent magnet synchronous machines (PMSMs), because of the influence of environment, magnetic saturation and so on, the electrical parameters are inconstant and thus it is critical to accurately identify these parameters. This paper proposes an effective PMSMs parameter estimation method to improve the identification accuracy and anti-interference ability. Specifically, this paper considers the influence of the voltage-source inverter (VSI) and magnetic saturation fitting polynomial in the PMSMs model. Then, in order to reduce the coupling interference of the parameters, this paper proposes a multi-parameter decoupling method based on adaptive genetic algorithm(AGA) for accurate parameter estimation. It decouples the PMSMs model into three parts by different speed to identify resistance, inductance and flux linkage separately. Compared with the existing methods, this approach can effectively identify the parameter in the PMSM model, which is verified experimentally on a laboratory PMSM.
What problem does this paper attempt to address?