LAUSR

This paper presents an improved model-based predictive direct torque control (MPDTC) to improve torque accuracy and reduce torque ripples which is a major issue in conventional direct torque control (DTC). Hysteresis controllers and traditional DTC switching tables are replaced by a model predictive controller to achieve an online optimization for voltage space vector selection and optimal duty ratio modulation method for torque ripple reduction. In order to provide an accurate motor model for MPDTC, novel offline and online motor parameter estimation methods are proposed to improve performance of the proposed MPDTC. The proposed parameter estimation adopts Popov’s hyper stability theorem to estimate accurate motor parameters, such as stator resistance, stator inductance and rotor flux linkage, which are critical for torque and flux estimation. The parameter adaptive MPDTC is verified by a hardware in the loop emulation platform, and experiment result is demonstrated using a dynamometer test bench, which therefore proves the feasibility of the proposed method.