Notably， predictive control is an optimal control strategy that reduces the impact of system delays by predicting and adjusting the future behavior of a motor using mathematical models. System delays are nonlinear time-varying disturbances that are often difficult to accurately estimate and compensate for. Moreover， they degrade the position tracking performance of conventional position predictive control when applied to fast responding objects such as linear motors. This study analyzes the sources and effects of system delays in a permanent magnet synchronous linear motor. A multi-step virtual positive predictive control is proposed by introducing an active variable speed coefficient into the prediction model to reduce the system delay and speed up motor response. This addresses the observed decrease in motor speed with a decrease in the position prediction error during the tracking process of conventional predictive control， which often leads to the failure of rolling optimization， and realizes time delay compensation for high-performance systems. This study experimentally simulates a low-power permanent magnet linear synchronous motor for medical microscopes. The proposed method can accelerate the dynamic response of the motor and reduce the effect of system delays on the tracking of different types of curves such as triangular and sinusoidal. In particular， the position tracking accuracy is improved by approximately 20% over that of conventional position predictive control.