电芯卷绕过程中，正负极片的头尾部在卷绕起始与结束时处于悬空、无张力的状态。受卷绕设备结构限制，卷绕时极片头尾部各有约200 mm长的悬空段，该段极易发生偏移。且来料自身质量越差，偏移程度越严重，变化趋势越明显。本文基于卷绕过程中极片的初始态检测、变量检测、结果反馈及条件补偿四个方面预测极片卷绕时头尾部偏移量，并实现闭环控制，从而提升电芯质量，并达成无人化、智能化控制的目的。

During the battery cell winding process, the head and tail of the anode and cathode are in a suspended, tension-free state at both the beginning and end of winding. Due to the structural influence of the winding equipment, the lengths of the electrode head and tail sections in this uncontrolled state can reach approximately 200 mm, leading to significant offset risks. The poorer the quality of the electrode material itself, the more severe the offset magnitude and variation trend become. This paper proposes a method to predict the head-tail offset during winding and achieve closed-loop control through four key aspects: initial-state detection, variable monitoring, result feedback, and conditional compensation. The proposed approach aims to enhance cell quality while enabling unmanned and intelligent control of the winding process.