智能制造工厂光控数字空间技术与基础理论研究

Nowadays, intelligent manufacturing has been the research cutting edge of manufacturing informatization, however, its technology and method of concrete implementation are still under study. Cyber-physical system (CPS) is an important form to realize the intelligent manufacturing which is a valid architecture that connects the manufacturing entity and the information space, but its underlying perception technology of the geometric data in the shop floor remains a further deep research. This project proposes an idea, the light-controller digital space, in the viewpoint of integration of the light-controller measuring methods and the applications of measuring data. The measuring requirements of geometric data in the shop floor are analyzed, the configuration selection and the method of layout optimization for the measuring instruments are studied, and the construction of hardware platform of light-controller digital space for the shop floor is realized. The uncertainty fusion methods of the diverse-source measuring data are investigated, the construction method of the storage model and theirs hierarchical relations are explored, and the self-adaptive data fusion and inference method based on the dynamic measuring load are studied. Moreover, the transformation tools of measuring data and the construction method of arbitrary local frames in the whole space around the shop floor are explored, which enable the integrated packaging and interface of the information for the light-controller digital space, and guarantee the convenient, real-time and efficient extraction and application of the geometric data in the shop floor. In all, it provides a support of the elementary theory and enabling technology for the method research of data perception in the intelligent shop floor.

智能制造已经成为当今制造业信息化的研究前沿，然而其具体实现技术与方式依然在探索中。信息物理系统作为智能制造实现一种重要形式，是联系制造实体与信息空间的一种有效架构，但其底层的制造现场几何量数据感知技术依然有待进一步深入研究。本项目面向智能制造现场的几何量测量需求，从光控测量方法集成与测量数据的有效整合及应用角度提出了光控数字空间这一理念。剖析车间层几何量的测量需求，研究测量仪器的选型配置与布局优化方法，形成车间层光控数字空间硬件载体构建方法。研究异源测量数据的不确定融合方法，探索测量大数据的存储与层次关系模型构建方法，研究基于动态测量负载的自适应数据融合与推理生成方法，寻求车间全域空间中任意位姿局部坐标系的构建方法与数据变换工具，实现了光控数字空间数据信息的整合封装与接口使能，确保了车间全域几何量的便捷、实时、高效提取与应用，为智能制造现场的数据感知方法研究提供了基础理论与使能技术支撑。

智能制造车间全域内各类几何量的精准测量与感知是打通制造现场与高维制造状态建模优化调度的关键，是智能制造推进实施的重要保障与研究前沿。本项目针对车间全域内几何量的精准感知，提出了光控数字空间这一理念，以分布式异构测量网络为主要载体，在车间级测量场的构建与使能方面开展了理论与关键技术的探索。研究了异构测量场的基础内外参数标定模型，进而提出了内外参数联合的标定方法，及针对车间全域尺度的基于多目视觉的异构测量系统定向方法，形成了精确的车间域测量场精准定向理论；研究了异构测量系统的误差传递模型，提出了面向共面约束方程组的不确定度解析计算模型；构建了综合多站测量可达性与空间不确定度场约束的测量场优化模型，提出了面向物理边界约束下的测量场布局优化进化求解算法，并在某大型车间布站进行应用；提出了基于光脉冲编码的车间域传感器时钟同步机制，构建了基于光同步的R-LATs系统，并进一步基于空间时钟的一致性，研究了多目标并行动态测量的测量模型修正方法，形成了面向智能制造车间内几何量与时空域的同步精准测量机制；面向车间内分布式多测量需求，提出基于测量场网格划分的海量测量数据与通信负载分治模式，形成了将测量数据的采集、传输、可视化及应用派发至边缘端的模式，有效解决了测量任务去中心化，确保车间作业中几何量感知系统的运行轻量化与边界实施的服务应用接口实现。本项目的研究为智能制造车间CPS架构中现场几何量的感知与车间级运行维护提供了新的视角，提出的基础理论与关键技术已在航空航天等领域进行了初步应用验证，具有重要的学术与工程价值。