基于多尺度图谱的选区激光熔化成形精度的在线监测研究

As a new technology of laser additive manufacturing, Laser Selective Melting has extremely important value and broad application prospects in the field of high-performance complex parts manufacturing, such as aerospace, biomedicine, automobile and so on. It is one of the hot topics in metal material forming research. Unfortunately, the current Laser Selective Melting processes lack online monitoring with feedback and accurate quantitative process control, which lead to the cumulative effect of the forming quality defects, reduce the forming precision and mechanical properties of parts, and seriously hinders its industrial application. To solve these problems, the project combine the digital manufacturing characteristics of Selective Laser Melting forming to the forming mechanism, explore the quantitative relationship among pool temperature and dimensional changes and forming accuracy, and develop multi-scale based defect monitoring algorithm through the newly developed spectral analysis mathematical theory. Also, the spectral graph theory is used to learn the model parameters, and find out the variables affecting the forming accuracy, and then an on-line intelligent monitoring and control system for forming precision and layer-by-layer defect feedback is developed. The research of this project is one of the key technologies for the development of high-performance metal additive manufacturing equipment. It has important academic significance and application value for the automation, intelligence and industrial application of additive manufacturing technology.

作为一种新型的增材制造工艺，激光选区熔化成形技术在航空航天、生物医学、汽车等复杂零部件的制造领域具有重要的应用前景，是当前金属材料成形研究的热点之一。然而，当前的选区激光熔化成形工艺缺乏具有反馈的在线监测以及精确量化的过程控制，引起质量缺陷累积效应，降低了零件的成形精度与力学性能，严重阻碍了它的工业应用。针对这些不足，本项目将结合选区激光熔化成形的数字制造特点，研究成形机理，探索熔池温度和尺寸变化与成形精度之间的量化关系，通过最新发展的图谱分析数学理论，开发多尺度的缺陷监测算法。同时，利用图谱算法学习模型参数，发现影响成形精度的变量，并以此开发成形精度和逐层缺陷反馈的智能在线监测与控制系统。本项目的研究是研发高性能金属增材制造装备的关键技术之一，对增材制造技术的自动化、智能化以及工业应用有着重要的学术意义和应用价值。

成形零件的质量缺陷至今仍是阻碍选区激光熔化技术广泛应用的技术瓶颈，成形过程的在线监测技术是解决这类难题的重要途径。现有的成形精度监测系统缺乏系统性地考虑工艺过程与监测系统的结合，不能形成工艺参数与成形质量之间的精确量化关系，难以实现有效的精度监测。项目以选区激光熔化成形的熔池变化及其零件成形精度作为研究对象，开发了在线监测方法，研究了SLM成形机理与熔池温度变化的物理过程。采集熔池内熔融程度不同和液滴、粉末飞溅造成的激光反射光线产生扰动、异常变化信号，开发了缺陷检测算法、图像处理及多尺度分析方法提取熔池尺寸、面积以及运动变化信号，揭示了熔池变化-温度-信号变化之间的关系；利用深度学习算法学习成形表面形貌与成形工艺参数之间的映射关系，最终建立了成形尺寸变化及缺陷产生与成形工艺参数之间的量化关系；将缺陷引起的成形几何误差作为反馈，通过调整不同工艺参数进行离线训练得到最优工艺参数，实现逐层反馈的自适应控制。项目研发中获得了大量熔池变化实时监测数据，以及采用图像处理、深度学习等算法获得的分析数据。项目为解决通过熔池温度监测与控制选区激光熔化过程的关键难题做出了重要贡献，为进行选区激光熔化成形的状态监测提供理论基础，对选区激光熔化增材制造技术的自动化及智能化提升有着十分重要的意义