多并联共融机器人大型光学镜面加工装备动态耦合特性及协同控制策略研究

With the development of modern optical system towards the direction of large aperture, high precision, high resolution and high power, the optical manufacturing equipment with multi manipulators has received attention increasingly. The manufacturing precision and efficiency have a great influence on the imaging quality and stability of optical system. Therefore, it is more urgent to study the dynamic coupling characteristics of co-fusion robot with multi parallel manipulators and propose the coordinated control strategy in such a difficult situation with restriction of working environment and technology of large optical mirror manufacturing and lack of ability in distributed autonomous and cooperative work of multi manipulators. The project seeks to discover the influence law of the dynamic coupling characteristics on the machining precision and master the differential evaluation method of the mirror surface machining precision by establishing the analysis model of the dynamic coupling characteristics. The disturbance control method is proposed and the problem of the disturbance compensation is solved by analyzing the multiple disturbance factors and revealing the disturbance mechanism. The reasonable planning of the machining task of the Co-fusion robot with multi parallel manipulators is realized by establishing the distributed optimization decision model of cooperative task planning. A master-slave distributed coordinated control system of the Co-fusion robot with multi parallel manipulators is ultimately established by introducing the stability analysis theory and studying the optimal coordination control method. The large optical mirror manufacturing test system with multi parallel manipulators, which is the basis of the experimental verification, is optimized and improved. The research provides a theoretical and experimental basis for the research of the Co-fusion robot on the key technologies of multi manipulators cooperation and swarm intelligence control. Meanwhile, it provides scientific support for the application of the Co-fusion robot in large optical mirror manufacturing.

随着现代光学系统朝着大口径、高精度及高分辨率的方向发展，多机器人光学加工装备日益受到重视，其加工精度对光学系统成像质量及稳定性有很大影响。大型镜面加工受环境及工艺等环节约束，且多并联共融机器人加工装备分布自主及协同作业能力欠缺，研究其动态耦合特性，并提出协同控制策略变得愈发迫切。本项目试图通过构建多并联共融机器人动态耦合特性分析模型，发现其对加工精度的影响规律，掌握镜面加工精度差异化评判方法；分析多重扰动因素，揭示扰动机理，并提出扰动控制方法，进而解决扰动补偿问题；建立协同任务规划的分布式优化决策模型，实现多并联共融机器人加工任务合理规划；引入稳定性理论，研究其最优协同控制方法，构建多并联共融机器人主从分布式协同控制系统；优化并完善试验系统，且进行实验验证，为共融机器人在多机器人协作与群体智能控制等关键技术研究奠定理论及实验基础，同时也为共融机器人应用于大型光学镜面加工提供科学支撑。

为适应现代光学系统大口径、高精度及高分辨率的发展要求，基于共融机器人基础理论与关键技术，设计研制了多并联共融机器人大型光学镜面加工装备。开展了光学镜面主动分区柔性液压支撑机器人系统设计、动态扰动及误差分析，五自由度混联机器人系统设计及运动学、动力学、惯量匹配特性、刚度特性、稳定性、精度特性、运动副摩擦补偿及人-机-环境信息融合与控制等系列研究工作。考虑光学镜面加工扰动，建立了柔性支撑系统动态扰动模型，分析了支撑系统性能参数与环境影响参数之间的映射规律；基于设计的五自由度混联机器人系统，分析了机器人运动学性能和动力学性能并构建了机器人的惯量匹配模型；建立了机器人的静刚度模型，并进行了虚拟样机验证；分析了机器人在其工作空间的耦合度变化情况，研究了研磨系统对机器人耦合特性的影响；基于李雅普诺夫模型，分析了机器人的稳定性；构建了机器人几何误差模型，并利用粒子群优化算法对机构误差源进行综合优化；考虑运动副的摩擦和碰撞，设计了基于前馈PD控制的关节摩擦补偿控制器；基于遗传算法改进的BP神经网络构建了考虑人-机-环境的机器人误差补偿策略；搭建了光学镜面主动分区柔性液压支撑机器人和五自由度混联加工机器人系统共融的大型光学镜面加工装备，并进行了实验研究。研究成果为多并联共融机器人在大型光学镜面加工方面提供了理论及实验基础。