부산대학교 도서관

The periodic motion of a permanent magnet linear synchronous motor (PMLSM) used in industrial applications requires considerable acceleration and deceleration at the start–stop stage, resulting in significant peak position errors. Moreover, thrust fluctuation due to the cogging effect, end effect, armature reaction, asymmetry of winding parameters, and time-varying electrical parameters will degrade the position-tracking precision at constant velocity. Thus, to avoid these impacts on position control accuracy, this article proposes an improved anticipatory fractional-order iterative learning control (ILC), named the anticipatory lead fractional-order ILC (ALFOILC). ALFOILC is superior to the anticipatory $\rm {D}^\alpha$-type fractional-order ILC (ADFOILC) regarding amplitude and phase compensations. Thus, ALFOILC increases the learnable band further. Moreover, this work analyzes the design, approximation, discretization, and tuning strategy for the decoupling parameters associated with ALFOILC. Finally, we build an experimental platform with an air-bearing PMLSM to validate the efficacy of ALFOILC through comparative experiments against ADFOILC and integer-order ILC.