Inductance Estimation of PMSM Using Extended Kalman Filter

Abstract

Estimation of parameters of Permanent Magnet Synchronous Motor (PMSM) plays an important role for motor controller tuning in Electric Vehicle (EV) application. Under running condition motor parameters vary due to different effects such as temperature, saturation and Voltage Source Inverter (VSI) non-linearities. Identification of parameters in running condition increases the control performance of system. This paper uses Extended Kalman Filter (EKF), which allows estimation of d-q axis inductances of PMSM. The control algorithm considered here is Field Oriented Control (FOC) for EV system having position sensor. The simulation is performed using MATLAB- Simulink software. The simulation results show that EKF identifies the d- and q- axis inductances Ld, Lq considering measurement and process noise using a state space model of motor equations for implementation.

Identification of Parameters and States in PMSMs

“…Table 3 reports the results from using the method proposed in [21] with the help of a PI controller.…”

“…In Figure 4, the identification of the flux represented over time using the PI approach is shown, see [21], with the help of Matlab/Simulink simulations.…”

“…Table 3 reports the results from using the method proposed in [21] with the help of a PI controller. In Figure 4, the identification of the flux represented over time using the PI approach is shown, see [21], with the help of Matlab/Simulink simulations.…”

“…A PMSM was modeled in the simulation with a nominal stator resistance of 0.1 Ω. After that, a proportional-integral (PI) controller with a fixed gain of 1.0 was used to operate the machine [21]. According to the simulation results, the motor can produce a torque of 10 Nm and a steady-state speed of 3000 RPM.…”

“…The motor was controlled using a proportional-integral (PI) controller with a fixed gain of 1.0. The simulation results are presented in Table 1 below: As seen in Table 1, which reports the results from using the method proposed in [21] with the help of a PI controller, when the stator resistance was artificially increased to 0.2 Ω, the motor was not able to achieve the desired speed and torque. The controller was also unable to maintain stable operation, resulting in poor performance and efficiency.…”