A fast online full parameter estimation of a PMSM with sinusoidal signal injection

In this paper, a fast online parameter estimation scheme with sinusoidal d-axis current injection is introduced. Compared to the existing parameter estimation strategies, the proposed estimation scheme has relative faster convergence time and is feasible in the transient operation. This estimation scheme is realized by the recursive least square algorithm based on different operating points of the Permanent Magnet Synchronous Machine (PMSM). With the proposed estimation scheme, the four parameters of the PMSM can be simultaneously estimated online. The derivatives of the d- and q-axis currents are considered in the proposed estimation scheme so that it is feasible in both steady state and transient state operation. The sampling principle is discussed to avoid the rank deficient problem for the estimation. An averaged sliding window is introduced for the estimation to minimize the estimation error from the A/D conversion and to reduce the influence of the measurement noise. The proposed estimation scheme is validated by the experimental results and compared with the parameter estimation with square wave current injection.