ARX Structure Based Predictive Torque Control of IMD with Improved Flux Estimation

The accuracy of the conventional predictive torque control (CPTC) of an induction machine relies on the precise machine parameters and combinatory estimation of these parameters is tedious since they are coupled and continuously varying with operating conditions. To make control robust against imprecise model and variations in machine parameters, autoregressive exogeneous structure-based predictive torque control (ARX-PTC) is presented in this work. ARX structure is formed by relating discrete time input and output samples using a linear difference equation, in which coefficients are tuned online with the recursive least square (RLS) method. Adaptive gain enhanced second order generalized integrator (AG-SOGI) along with the gain normalized frequency locked loop (FLL) is employed for flux estimation that nullifies the issues of DC drift, harmonics and disturbances present in sensed signals. Since the gains of AG-SOGI are tuned by normalized speed error, it gives better filtering during steady state and enhanced performance during dynamics. The presented control algorithm is simulated in Matlab/Simulink platform and the effectiveness of the control is verified from the observed results.