本论文围绕复杂曲面零件高效高质量加工的应用需求，以一种五轴全并联加工机器人为研究对象，以实现机器人高速高精运动控制为目标，系统地开展了驱动参数优化、刀具路径规划、进给速度规划、轨迹跟踪控制四方面研究。具体研究内容如下：针对多指标耦合作用下的加工机器人优化难题，根据多种工作模式对机器人性能的需求，提出了设计指标体系分解策略，建立了基于运动轨迹的五轴并联加工机器人驱动参数优化方法，实现了驱动单元与机器人应用需求间的性能匹配，提升了机器人动态性能。根据五轴全并联机构构型与驱动参数优化结果，搭建了高灵活性高动态特性并联加工机器人样机。针对并联机器人多轴耦合运动姿态调整灵活性高的特点，提出了样条曲线间连续性约束条件以及并联机器人刀轴矢量规划策略，建立了高阶连续与调姿灵活的刀具路径规划方法，通过三轴和五轴刀路规划与加工实验，证明了本论文提出的规划方法可以克服传统方法在刀具通过极点姿态时面临的刀路失真问题，充分发挥出并联机构姿态耦合运动的优势，提升了姿态调整精度与效率。建立了并联机器人驱动轴跟踪误差预估模型，揭示了非线性动力学和摩擦耦合作用下机器人跟踪误差产生机理，提出了一种结合动力学前馈与进给速度规划的综合控制方法，解决了多轴耦合运动学特性以及非线性动力学特性影响下并联机器人的高精度控制难题，通过刀轨跟踪和加工实验验证了提出的控制方法在提升加工效率和加工质量方面的有效性。根据跟踪误差预估模型，提出了驱动轴跟踪误差预补偿、末端轮廓误差估计与预补偿两种轨迹跟踪控制方法，开展仿真实验对比分析两种控制方法的性能，采用对加工轨迹连续性影响较小的轮廓误差预补偿方法进行了刀轨跟踪实验，证明了提出的轨迹跟踪控制方法可以保证机器人跟踪大曲率或曲率激变曲线路径时的控制精度。 基于以上理论体系与关键技术，实现了具有复杂曲面特征的飞机框架类薄壁结构件与S试件的高效高质量加工。本论文研究工作对开发高性能五轴并联加工机器人以及促进其在工业中的推广应用具有重要意义。

Focusing on the high-efficiency and high-quality machining of complex curved surface parts, this paper takes a 5-DoF fully parallel machining robot as the research object and aims to realize high-speed and high-precision motion control. Systematical research about the driving parameter optimization, toolpath planning, feedrate scheduling and trajectory tracking control is conducted. The details are as follows:For the optimization design of parallel robots under multi-index, a design index system decomposition strategy is proposed based on the performance requirements of the real work mode. The motion trajectory-based driving system parameters optimization method for five-axis parallel machining robot is further established. The match between the driving unit and the performance requirements is achieved, in this way, the robot’s dynamic performance is improved. Based on the 5-DoF fullly parallel mechanism and the driving parameters optimization results, a parallel machining robot prototype with high flexibility and dynamic characteristics is built.Aiming at the advantage of high attitude adjustment flexibility, the continuity constraints between spline curves and the the parallel robot tool axis vector planning strategy are proposed, and a high-order continuous and attitude adjustment flexible toolpath planning method is established. Through the three-axis and five-axis toolpath planning and machining experiments, it is verified that the proposed toolpath planning method can take full use of the advantages of the attitude coupling motion property, and improve the accuracy and efficiency of the attitude adjustment.The driving axis tracking error prediction model is established, and the tracking error generation mechanism under the coupling action of the nonlinear dynamics and friction is disclosed. On the basis, an integrated control method that combines the dynamic feedforward and feedrate scheduling is proposed. The proposed method overcomes the challenge of parallel machining robot’s control precision guarantee under the effect of multiple axes coupling motion and nonlinear dynamics characteristics. The effectiveness of the proposed control method in improving machining efficiency and quality is verified through toolpath tracking and machining experiments.According to the tracking error prediction model, two trajectory tracking control methods including the driving axis tracking error precompensation, tool contour error estimation and compensation are established. Simulations are conducted to compare the performance of the two proposed control methods. The contour error compensation method is adopted to carry out the trajectory tracking experiments. The experimental results illustrate that the proposed control method can guarantee the robot’s control precision when tracking toolpaths which have the characteristic of large curvature or sudden changing curvature.Based on the above theoretical system and key technologies, the high-efficiency and high-quality machining of the S-shaped test part and the frame-type structural part is achieved. The research work in this paper is of great significance to the development of high-performance 5-axis parallel machining robot and the promotion of its application in industry.