| Summary: | 碩士 === 國立成功大學 === 工程科學系 === 87 === In recent years, several methods for a field-oriented induction motor control without rotational transducers have been proposed. These methods have disadvantages in that the load torque disturbance and system parameter variations cause speed estimation errors in the induction motor. Therefore, for a sensor-less induction motor speed control, simultaneous estimation of motor speed, rotor flux, load torque and rotor time constant is required. This thesis presents an application of the extended Kalman filter (EKF) to a sensor-less induction motor speed control. The proposed method can estimate the motor speed, rotor flux, load torque and rotor time constant simultaneously by measuring the stator currents and voltages. The EKF has been known to be capable of estimating system parameters and state variables correctly by eliminating the influences of structural noises.
First, this thesis proposes a full order induction motor model as the EKF model for a sensor-less induction motor speed control. Second, because the stator current can be measured using the current sensor, a reduced order EKF model for a sensor-less induction motor speed control to reduce the computing time is presented. The EKF method can accurately estimate the motor speed, the load disturbance and the system parameters. Therefore, the proposed method is suitable for sensor-less speed control and is robust against the load torque disturbance and system parameter variations.
The proposed control scheme was implemented using a 32-bit TMS320C32 microprocessor. Simulations and experimental results demonstrate that the proposed control scheme can rapidly and accurately estimate the motor speed, the load disturbance, the system parameters and has a robust speed response. Hence, according to the simulation and experimental results, agreement with the theoretical basis of the proposed scheme has been demonstrated.
|
|---|
