mplémentation d'une Identification En Temps Réel De La Machine Asynchrone A Cage Sur Le DSP TMS320 LF 2404 A

Abstract

This work is for objective the development of identification and estimation technics of a state which are not directly measured by linear Kalman filtering and extended Kalman filtering of the asynchronous machine. The identification algorithms by Kalman filter applied to the induction machine are tested for validation by simulation under MATLAB SIMULINK environment. Around a development system based on digital signal processor " DSP TMS 320 LF 2407 EVM " of Texas Instrument, for application to monitoring and control of electric machines, we have developed a test bench for study and integration of real time monitoring algorithms of induction machine. The device including in addition to the asynchronous machine, two electronics modules for filtering and adjustment of measurable signals and a " L100" voltage inverter of HITACHI. The real time estimation algorithm of the rotor state (rotor flux) of the induction machine was successfully developed on this device. Key words: identification, real time, induction motor, linear Kalman filter, extended Kalman filter, signal processing processor TMS320LF2407.