轮廓跟踪控制是当前精密机电控制领域内的一个重要研究方向。本文以提升轮廓跟踪精度为研究目标，立足于内模原理控制方法，通过内模原理使能的渐近跟踪特性，确保轮廓误差满足现代制造业中的超高精度需求。进一步，本文基于双轴轮廓、多轴轮廓等不同应用场景，针对性地设计具有不同形式的单轴内模控制器及多轴轮廓控制策略，以完成多种工况下的精密轮廓跟踪控制任务。本文的主要研究内容如下： 面向双轴轮廓跟踪控制，设计了一种基于单轴重复控制与多轴交叉耦合控制的复合控制方法（RC-CCC），对于周期轮廓信号可实现渐近跟踪。基于重复控制的结构特点，选择了正确的轮廓误差反馈位置。随后，基于所提出的RC-CCC方法的单轴RC控制器展开鲁棒改进设计。本文通过仿真、运动控制实验以及微立体光固化加工实验，验证了该控制方法的可行性及有效性，可加工特征尺寸为2.5μm的光固化样件。 重点面向多轴轮廓跟踪控制展开研究，设计了一种基于位置主元主从结构的时变内模轮廓跟踪控制方法（TV-IMCC）。拓展了位置主元框架，通过一种信号变换算法，增加了TV-IMCC可跟踪的轮廓种类。针对位置域建模引入的时变参量，设计从动轴时变内模控制器及时变镇定器，从而实现整体多轴轮廓的渐近跟踪。一系列仿真与实验结果验证了所提出的TV-IMCC方法相比现有方法更适用于多轴轮廓跟踪，且对于尺寸约30mm的轮廓参考信号，可实现约60nm的轮廓跟踪误差。 进一步，基于所提出的TV-IMCC方法的单轴时变内模控制器展开鲁棒改进设计（Robust TV-IMPC）。通过设计灰盒ESO观测器，对系统中存在的模型不确定性及外界扰动进行反馈补偿，从而将实际被控对象模型转化为带有加不确定性的标称模型。仿真与实验结果表明，经过鲁棒改进设计后的TV-IMPC方法的跟踪精度相比现有时变内模方法能够进一步提升，对于行程大于80mm的参考信号，其跟踪相对误差可达1e-6量级。 最后，将所提出的TV-IMCC方法应用于振镜-伺服运动台协同激光加工系统中。为使能协同激光加工，研究并设计了一种基于五次多项式路径平滑的轨迹分配方法。基于该轨迹分配方法与TV-IMCC控制方法，成功实现了振镜-伺服台协同激光加工，并获得了整体尺寸160mm，特征尺寸<1mm，相对误差<3e-4的加工图案。

Contouring control is an important branch of the precision mechatronic servo control. In the contouring control, the axial controllers and the contouring control methods work simultaneously to enhance the contour tracking precision. The contouring control is widely used in many advanced equipment manufacturing fields, such as advanced machine tools, lithography machining process, micro-/nano-scale machining and inspection, etc. With the increasing demand for contouring precision, the limitations of the existing contouring methods are revealed: for two-axis contouring, the axial controller lacks of asymptotic tracking precision; for advanced multi-axis contouring, both high-precision axial controllers and multi-axis contouring methodologies are under research. To solve the above open problem, this paper designs contouring methods based on the internal model principle (IMP). Thanks to the asymptotic tracking performance enabled by the IMP, the contouring errors are able to satisfy the industrial requirements. Specifically, different kinds of the IMP-based contouring controllers are designed in this paper for different working conditions, including the two-axis and multi-axis contouring methodologies. The main results of this paper are listed as follows: Firstly, an RC-CCC control methodology is proposed for two-axis contouring, where the axial RC controller and the CCC method are combined. The RC-CCC method is capable of asymptotically tracking periodic signals. Based on the specific structure of the RC, The correct feedback position of the contour errors is determined in this paper. The system stability and tracking performance are analyzed in the equivalent CCC system. Various simulation and experimental results validate that the proposed RC-CCC is able to obtain contour errors of <100nm when tracking references of 100μm amplitude. Moreover, the robustness of the axial controller of the RC-CCC is improved. Specifically, the axial controller is replaced with an RBF neural network-based parameter adaptive controller. As a result, the system is more robust under unknown system nonlinearities. Moreover, the plug-in RC structure is integrated with the robust controller, so that the proposed method is able to track periodic references in high precision. Various simulations and experiments based on a micro-stereo-lithography (MSL) system are conducted to validate the feasibility and effectiveness of the proposed method, and MSL workpieces with feature size of 2.5μm are obtained. Secondly, a mater-slave position domain-based time-varying internal model contouring control (TV-IMCC) is proposed for multi-axis contour tracking. The multiple axes are divided into master and slave ones, and the slave axis model dynamics is constructed in the master axis position domain, so that the initial multi-axis contouring system is decoupled into two-axis subsystems. The position domain framework is extended as well. Specifically, the categories of contours to be tracked are extended by designing a signal converting algorithm. A time-varying system is obtained after transforming the position domain exosystem into the time domain. Then, the time-varying internal model controller and stabilizer are designed to finally enable multi-axis asymptotic contouring. Abundant simulation and experimental results validate the multi-axis contouring capability and high contouring precision of the proposed TV-IMCC compared to the existing methods. For contours with 30mm stroke, the error of the TV-IMCC is about 60nm. Furthermore, the robustness of the time-varying internal model principle-based control (Robust TV-IMPC) is guaranteed. A gray-box ESO is designed to compensate the model uncertainties and external disturbances of the system, so that the model of the plant can be transformed to a nominal one with a small additive uncertainty. The corresponding time-varying internal model controllers and stabilizer are designed for this plant model. Various simulations and experiments are conducted, indicating that the proposed robust TV-IMPC can increase the tracking precision compared to the existing method. The relative error of the robust TV-IMPC is of 1e-6 with a stroke larger than 80mm. Finally, the TV-IMCC methodology is applied to a scanner-stage synchronized laser manufacturing system. To enable the fabrication, a trajectory distribution method is designed by quintic polynomial path smoothing. As a result, by solving a standard quadratic optimization problem, the general contour trajectory can be distributed into axial trajectories, and the motion constraints for each axis are satisfied as well. The laser manufacturing is successfully conducted with the proposed TV-IMCC and trajectory distribution method. Workpieces with general size of 160mm, feature size of <1mm and relative error of less than 3e-4 are obtained.