Direct Flux Vector Control of Synchronous Motor Drives: A Small-Signal Model for Optimal Reference Generation
Anantaram Varatharajan, Gianmario Pellegrino, Eric Armando
Abstract
A novel direct flux vector control (DFVC) scheme is presented based on the real time use of the motor small-signal model for optimal reference generation without preprocessed look-up tables (LUTs). The control scheme is valid for reluctance- and PM-synchronous machines. The stator flux magnitude and the load angle are the controlled variables and the optimal reference values respecting maximum torque per ampere (MTPA), maximum torque per volts (MTPV), voltage, and current limit conditions are computed in real time from the small-signal model. Analytical expressions of MTPA and MTPV criteria are derived to enable online adaptation according to the small-signal approximation of the motor model. The motor parameters reside in the flux-map LUTs used in the flux observer; besides that, no additional tables are necessary. Furthermore, online parameter adaptation is proposed to further improve torque tracking accuracy against flux-map LUTs errors. The feasibility of the proposed scheme is demonstrated through experiments on a 1.1-kW synchronous reluctance (SyR) machine test-bench. The proposed control scheme simplifies the implementation and calibration of the DFVC, while improving its MTPV control and its roughness against model parameter errors. Prospective fields of application are spindle and traction drives.