When a permanent magnet synchronous linear motor performs a tracking control task with repetitive characteristics, the position error of the system also changes periodically. Aimed at this problem, a control strategy based on offline iterative learning compensation to optimize the linear motor position input signal was proposed. The interference of the linear motor system was carried out through offline iterative learning. Iterative learning of camshaft trajectory and sin curve trajectory was carried out to realize the tracking control of the permanent magnet linear motor motion trajectory and reduce the tracking error of the linear motor. The experimental results show that the position input signal following error generated by off line iterative learning control is small, and the position error caused by tracking lag is well eliminated.