Accurate Discrete-Time Modeling for Improved Torque Control Accuracy for Induction Machine Drives at Very Low Sampling-to-Fundamental Frequency Ratios

Operations at low sampling-to-fundamental frequency (S2F) ratios of induction machines are required for high-power or high-speed applications. The traditional indirect field-oriented control is open-loop control on torque, which is sensitive to machine parameters and degrades at low S2F ratios. Deadbeat-direct torque and flux control (DB-DTFC) is closed-loop control on torque but also degrades at low S2F ratios with inaccurate discrete-time approximations. This article presents a methodology of accurate discrete-time modeling for the flux and current observer and the deadbeat controller used in DB-DTFC. The current, flux, and torque estimations are improved with the proposed observer model at very low S2F ratios. The torque control is enhanced and insensitive to machine parameters under high-speed low-S2F-ratio conditions. The increased computational time of the proposed model is insignificant for real-time implementations.