单面焊双面成形熔孔动态行为及背面成形质量智能控制

Technology of fusion keyhole one-side welding with back formation is very applicable to butt weld with gap under the situation that the backside of work piece is unreachable. The fusion keyhole is benefit for good quality of back weld appearance. But the equipment used to realizing automation of the technology requires higher consistency of gap and the robust of equipment is poor. In this project, a monitor and control system is established based on visual sensor. Fusion keyhole, weld pool and gap are detected simultaneously. The Influences of welding conditions on morphological parameters of fusion keyhole are analyzed and the dynamic behavior of fusion keyhole is researched. Based on monitoring result, numerical simulation model is founded. Formation mechanism of fusion keyhole is revealed and the conditions to form stable fusion keyhole are also studied. The relationship among fusion keyhole, weld pool, gap and quality of back weld appearance are explained. Based on the numerical simulation result, the model among welding parameters and gap, morphological parameters of fusion keyhole and weld pool and quality parameters of back weld appearance are identified By Particle Swarm Optimization. The intelligent control algorithm is put forward to realize the intelligent control of welding process. So, the works will be promote the progress of automation and intelligentization of fusion keyhole one-side welding with back formation.

熔孔型单面焊双面成形工艺适用于背面不可达、留有间隙对接焊过程，熔孔的存在利于获得较好的焊缝背面成形质量，但现有的单面焊双面成形自动化装备对间隙的一致性要求较高，抗干扰能力差。本项目构建视觉监控系统，采用视觉传感器同时检测熔孔、熔池和间隙，研究焊接工艺条件对熔孔形态的影响规律，分析熔孔的动态行为，采用数值模拟手段，探究熔孔的形成机理并分析获得稳定熔孔的条件，揭示熔孔与熔池的相互作用机制，挖掘熔孔、熔池及间隙与背面成形质量的关系，在此基础上，确定“焊接工艺条件-熔孔及熔池形态-背面成形质量”之间的物理模型结构，并采用粒子群优化算法辨识模型参数，设计熔孔型单面焊双面成形工艺的智能化调控方案，推动熔孔型单面焊双面成形工艺的自动化和智能化。

在背面焊接空间受到限制的工作环境以及重要的薄板焊接结构，需要单面焊双面成形技术。熔孔型单面焊双面成形工艺较渗透型更易于获得良好的焊接质量，在生产中的应用更广泛。但目前该工艺基本上都是由焊工手工完成，亟需实现该工艺的焊接自动化。.基于视觉检测，从工件正面的焊接区信息中，获取表征熔透等焊接质量的重要信息，进而实现熔孔型单面焊双面成形工艺的自动控制，对于推动焊接制造的自动化，具有重要的理论意义和现实意义。.本项目搭建了基于双视觉传感器TIG单面焊双面成形多信息监控实验平台，从试件正面和背面拍摄到了不锈钢、碳钢在不同焊接工艺条件下清晰的熔池、熔孔、间隙图像。针对拍摄到的熔孔、熔池、间隙图像，针对性的设计了包含滤波、锐化、形态学算法、Canny边缘检测算子等图像处理算法，提取了熔孔、熔池、间隙、熔孔区电弧长度等的特征参数。.建立了预留间隙对接TIG焊三维瞬态熔池传质传热模型，采用VOF法追踪熔池自由表面与熔孔界面，数值计算了温度场、熔池流场与熔孔行为的演变过程，分析了熔孔形成过程及其对熔池传质传热的影响。发现熔孔的出现加剧了熔池内部流动，有利于液态金属充分过渡到工件背面，促进工件进一步熔透。.分析了熔孔失稳的过程，并将小孔失稳过程分为纵向失稳、横向失稳以及渐缩失稳三类，并给出了他们的特征。研究了焊接工艺参数对熔孔的影响规律，发现焊接电流对于增大熔孔起到较大的作用，焊接速度对于调控熔孔的尺寸具有一定的不确定性，较焊接电流、焊接速度，送丝速度更利于调控熔孔尺寸，实现熔孔的稳定。.搭建了基于深度学习算法和支持向量机两种智能建模方法，对预留间隙TIG焊接穿孔／熔透状态进行预测，均取得了较好的效果。基于支持向量机建立了有效预测熔孔失稳的模型，预测模型能够提前“预警”。. 建立了以送丝速度为调控量的控制器，在变间隙情况下，实现了焊接过程中全程穿透型熔孔的稳定存在，保证了工件的熔透。