Abstract
This paper proposes a compensation method for interpolation error of the maximum power control of a PMSM control system that generates current commands through a look-up table. A torque control system using a look-up table created through experiments has superior characteristics, such as control stability and torque accuracy, compared to a system that executes torque control via a linear controller based on modeling. However, it is impossible to generate information on all the currents for the output torque in the table. Therefore, because the data stored in the look-up table have a discrete characteristic, they are linearly interpolated to generate a current command for the torque command. However, the PMSM current trajectory is generally elliptical, which causes an error owing to linear interpolation, reducing the maximum output power. In particular, when the table data are insufficient, such as in the high-speed operation range, the reduced maximum output cannot be ignored. This paper proposes a compensation method for the interpolation error using two feedforward compensators and a PI controller, which was verified through experiments.
Highlights
The permanent magnet synchronous motor (PMSM) application range has gradually broadened in recent years through automotive motor control, producing higher output than conventional motors
This type of automotive motor control requires robust control based on the input parameters
This paper proposes an improved interpolation error compensation method
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. [16,17] provide a method for applying a counter-electromotive force table to compensate for the interpolation error These methods require additional experiments to create a counter-electromotive force table, they can significantly reduce the current reference memory and be robust to parameter variation. Since this kind of method has a slow response and the current commands are incorrect when the voltage error is small, additional compensation is required. One of the experiment results is when only the voltage feedforward controller is applied to the PI controller, and another one is when two proposed feedforward controllers are fully applied
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have