Improved EKF-based direct torque control at the start-up using constant switching frequency

Abstract

A sensorless extended Kalman filter (EKF) based direct torque control (DTC) with the use of a constant switching frequency control (CSFC) of induction machine is proposed in this research work. This paper is meant to investigate the dynamic performance for DTC-CSFC and DTC with a hysteresis torque controller by applying different speed step changes from three different low speed operations. The dynamic performance results show that the sensorless EKF-based DTC-CSFC is superior to the conventional EKF-based DTC when reference speed change is applied from 0 to 59 rad/s. Additionally, improvement in the reduction of torque ripples is achieved for all step changes of speed. The effectiveness of the proposed EKFbased DTC-CSFC is confirmed by experimental validation.

Simple Flux Regulation for Improving State Estimation at Very Low and Zero Speed of a Speed Sensorless Direct Torque Control of an Induction Motor

“…, , , The stochastic continuous time system in (16) -(18) must be expressed in the discrete form in order to fit with the structure of extended KF (EKF) as follows: (19) (20) (21) The linearization of (20) is performed around the current estimated state vector , by using the Taylor series expansion with higher order terms neglected, as given in (22). (22) Assuming the system noise and the measurement noise as white zero mean noises, the recursive EKF equations are obtained by the following:…”

“…Also in this work, a speed estimator, which is based on an extended Kalman filter (EKF), is employed. The EKF based estimator has high convergence rate and good disturbance rejection, which can handle the model uncertainties and the effect of unmeasured disturbances [21]. In general, EKF techniques currently employed for speed sensorless IM drives can be classified into two types: reduced order [22] and full order [23], [24] estimators.…”

Experimental Evaluation of Torque Performance of Voltage and Current Models using Measured Torque for Induction Motor Drives

“…Toliyat et al [10] have developed artificial neural networks (ANNs) in closed loop observer for estimating rotor resistance and mutual inductance. There is also a stochastic approach that uses EKF in estimating the variables of IM, such as speed, torque, and flux [11]- [14]. Using EKF-based observer, it is possible to estimate the unknown parameters of IM, taking into account the parameter variations and measurement noises, in a relatively short time interval [15]- [16].…”

Comparison of Estimated Torques Using Low Pass Filter and Extended Kalman Filter for Induction Motor Drives

“…There is also a stochastic approach that uses extended Kalman filter (EKF) in estimating the variables of an induction motor (IM), such as speed, torque, and flux [3]. Using EKF-based observer, it is possible to estimate the unknown parameters of IM, taking into account the parameter variations and measurement noises, in a relatively short time interval [11]- [16]. This paper investigates the real time calculation of torque using the estimated state variables based on the LPF filter and EKF and then compares them with simulated torques.…”