Implementation of extended Kalman filtering algorithm with improved flux estimator on TMS320F28335 processor for induction sensorless drive

This paper describes a real-time speed estimation of an induction motor (IM) based on the extended Kalman Filter (EKF). The algorithm was implemented on the TMS320F28335 digital signal processor (DSP), which is capable to perform complex mathematical transformations. The proposed vector control system was firstly modelled in MATLAB Simulink and then verified using Code Composer Studio (CCS). An improved rotor flux estimator is introduced, which makes the presented estimation technique similar to the MRAS-based estimators.