Loss minimization for dynamic load trajectories on induction machine drives without torque performance degradation

This work is to provide a systematic loss minimizing solution for general dynamic load profiles on induction machines, without compromising desired torque dynamics. The drives use deadbeat-direct torque and flux control (DB-DTFC), in which torque production and stator flux linkage are fully decoupled and the Volt-sec (incremental stator flux linkage) vectors are the manipulated inputs. Stator flux linkage, as a separate degree-of-freedom, can be controlled dynamically to minimize machine losses at the same time that torque follows the commanded trajectory. With properly formed loss models and accurately captured machine dynamics, dynamic programming is adopted to generate loss minimizing stator flux trajectories for given load trajectories. A sinusoidal approximation for cyclical loading profiles and typical servo velocity profiles with large torque steps, are used to evaluate the proposed solution. Two filter-based methods which have been developed specifically for cyclical loading and servo velocity trajectories, respectively, are compared with the proposed method with respect to energy saving capability. Compared to the existing filter-based solutions, both simulation and experimental results demonstrate the proposed approach achieves better energy savings during the standard industrial trajectories examined here.