Search

This paper presents a phase-compensated microstepping (PCMS) for permanent-magnet stepper motors in order to overcome the steady-state position error (SSPE) limitation of conventional microstepping without position feedback. An additional phase to compensate for the SSPE is injected into the desired currents for PCMS. Proportional–integral (PI) current feedback with feedforward is proposed in order to guarantee the desired currents for PCMS. Given a desired position error and a desired velocity, the additional phase is calculated in order to ensure the desired position error. In the proposed method, the structure of the PI current feedback with feedforward is not modified, and the additional phase can be obtained by a linear interpolation with an offline generated lookup table. Therefore, it can be easily applied to existing industrial applications without any costs for additional hardwares. The experimental results validate the effectiveness of the proposed method. [ABSTRACT FROM PUBLISHER]

The application of a disturbance observer (DOB) and a variable structure controller (VSC) to current control of microstepping for permanent-magnet stepper motors (PMSMs) is studied. This is a recent approach to the current regulation of PMSM that results in improvement of position tracking performance of microstepping using only current feedback. We designed the DOB to estimate back electromotive force (EMF) in the form of a high-pass filter. Since the PMSM is a two-input one-output minimum-phase nonlinear system, the asymptotic convergence of the currents to the desired currents guarantees the asymptotic convergence of the position to the desired position in microstepping. Therefore, in order to compensate for the estimation errors of back EMFs and to guarantee the desired currents required for microstepping, the VSC is designed with a state feedback controller. The current tracking error and the integral of the current tracking error are proven to asymptotically converge to zero. Since the proposed method does not require any transformations or the calculation of sinusoidal terms, its structure is simple. The least lower bound of the absolute steady-state position tracking error during a constant-velocity period is estimated in order to analyze the performance of the proposed method. Our experimental results validate the performance of the proposed method. [ABSTRACT FROM PUBLISHER]