PMSM parameters estimation based on dual STF-UKF algorithm

Hui Lin,Shuaishuai Lü
DOI: https://doi.org/10.3969/j.issn.1001-0505.2016.01.009
2016-01-01
Abstract:To solve the parameter identification problem of the permanent magnet synchronous mo-tor, the identification model of PMSM( permanent magnet synchronous motor) was analyzed.Re-garding the parameters as slow varied states along with time and considering the presence of the sys-tem noise and measure noise, an improved unscented Kalman filter( UKF) algorithm was proposed based on the strong tracking filter( STF) .This improved filter algorithm is able to estimate the pa-rameters of PMSM including stator resistance, permanent magnet flux linkage and q-axis and d-axis inductance.The stability of the proposed algorithm was discussed.In order to reduce the calculation consumption of the algorithm, the four parameters were divided into two parts and estimated by dual STF-UKF, respectively.Simulation results show that the improved UKF algorithm can estimate the parameters accurately under different PMSM operations.
What problem does this paper attempt to address?