The feed system is a typical mechatronics system, as well as motion control equipment, that serves as a fundamental component of CNC machine tools. Developing a precise mechatronic integrated model of the motion control of the feed system is a crucial step in achieving the digital twin and virtual machining. The extant literature extensively covers the dynamic modeling and identification of feed systems, the characterization and traceability of position deviations, and the study and optimization of motion control algorithms. Nevertheless, there is a deficiency in the analysis and optimization of the uncertainty pertaining to the motion control position of the feed system. This paper focuses on the positioning repeatability and considers the random position deviation that occurs during the movement of the feed system. The current research on positioning repeatability primarily concentrates on static factors such as geometric error and assembly error. Three primary causes of uncertainty in the motion control position of the feed system are identified: disturbance uncertainty, parameter uncertainty, and model uncertainty. Each component of uncertainty is thoroughly classified and studied based on the specific characteristics of the feed system. This study provides a comprehensive overview of the current data-driven control approaches used to address uncertainty in the feed system. It specifically focuses on the data-driven controller turning method, feedforward compensation method, and controller design method. The review concludes by discussing the latest research progress and limitations and suggests hopeful trends. Specifically, when conducting integrated modeling of the feed system, it is crucial to thoroughly account for the impact of stochastic processes. Additionally, the issue of switching between dynamic and static states in the dynamic model should be taken into consideration. Simultaneously, it is necessary to examine a more suitable approach that integrates the system model with the uncertainty characteristics to develop a more robust optimization strategy for controlling and efficiently managing the spread of position deviation in the feed system.

中文翻译：

---

机床进给驱动运动控制中的定位不确定性综述

进给系统是典型的机电一体化系统，也是运动控制设备，是数控机床的基本组成部分。开发进给系统运动控制的精确机电一体化模型是实现数字孪生和虚拟加工的关键一步。现有文献广泛涵盖进给系统的动态建模和识别、位置偏差的表征和可追溯性以及运动控制算法的研究和优化。然而，对于进给系统运动控制位置的不确定性的分析和优化还存在不足。本文重点研究定位重复精度，并考虑进给系统运动过程中出现的随机位置偏差。目前重复定位精度的研究主要集中在几何误差、装配误差等静态因素上。确定了进给系统运动控制位置不确定性的三个主要原因：扰动不确定性、参数不确定性和模型不确定性。根据进给系统的具体特征，对不确定性的每个组成部分进行了彻底的分类和研究。这项研究全面概述了当前用于解决饲料系统不确定性的数据驱动控制方法。它特别关注数据驱动的控制器转动方法、前馈补偿方法和控制器设计方法。该评论最后讨论了最新的研究进展和局限性，并提出了充满希望的趋势。具体来说，在对供给系统进行集成建模时，彻底考虑随机过程的影响至关重要。此外，还应考虑动态模型中动态和静态之间的切换问题。同时，有必要研究一种更合适的方法，将系统模型与不确定性特征相结合，以开发更鲁棒的优化策略来控制和有效管理进给系统中位置偏差的扩散。