ANN based Torque estimation and Control of induction Motor drive used in Chocolate Industry

Abstract

The paper focuses mainly on ANN based estimator (to select suitable IM according to load profile) and controller (as future work), and effort will be given to prove that this technique to estimate and track the torque accurately and fast as compared to latest available DSP estimator and other conventional controllers. With the experimental setup (vector control technique) available in lab (in which DSP TMSLF 2407 technology has been implemented for firing of IGBTs and estimation of parameters) set of values will be generated for the parameters vds vqs ids iqs θ. Approximately 66000 values will be taken for each parameter. They will feeded to two ANN based estimator and Te estimated from there will be compared with Te estimated from DSP control board and it will be shown that ANN estimation technique gives better performance as compared with DSP estimation technique. Also the various parameters taken will also be compared and cross – checked with a vector control simulation model available in MATLAB 6.5.