磁控旋转电弧堆焊自由曲面特征识别理论与关键技术研究

Rotating arc surfacing under magnetic field have advantages including increase of bead width, decrease of penetration and dilution rates, as well as improving the transient responsive performance of arc sensor system.There would be broad prospects and potential developments in the application of magnetic-control rotating arc surfacing technology to remanufacture..Feature recognition of the worn surface is a key technology for weld forming and seam tracking during surfacing process. However, it is a great challenge to achieve surface features recognition in surfacing due to the structural diversity, irregular surface shape and less-obvious bead structure of remanufacturing components. And the method on free-form surface feature recognition for surfacing had not been established. To break through the limitations of traditional seam tracking method based on arc sensing, this research proposes a thought that solve free-form surface feature recognition problems through the full use of various information during rotating arc surfacing under transverse magnetic field and the application of statistical learning methods to investigate mathematics representation of spatial pose recognition of arc sensor. Thorough research on basic theory, implement method and related scientific issues about free-form surface feature recognition for magnetic-control rotating arc surfacing will be pursued to achieve great progress on seam tracking for multi-pass surfacing in remanufacturing. The research results will provide significant basis for constructing a set of theoretic systems of surfacing under magnetic field and promoting surfacing technology using accurate weld formation control for remanufacture workpieces.

磁控旋转电弧堆焊具有增加焊道宽度、减小熔深和降低稀释率等优势，且易实现扫描半径在线可调的电弧高速旋转，有助于大幅提高电弧跟踪传感系统瞬态响应，在再制造领域具有广阔应用前景和发展潜力。.有效识别磨损表面特征是实现堆焊焊道成形及跟踪控制的关键和前提。再制造零部件的结构多样性、非规则表面形状以及焊道特征不明显等因素给自动化堆焊时的表面特征识别带来困难，目前尚未形成服务于再制造堆焊的自由曲面特征识别成熟方法。本项目提出充分利用和发掘外加横向磁场作用下电弧旋转焊接过程的多种信息，从基于统计学习方法探索电弧传感器空间姿态表征等途径突破传统电弧传感检测技术在自由曲面特征识别局限的思想，拟针对磁控旋转电弧堆焊自由曲面识别的基础理论、实现方法和若干关键科学问题开展深入研究，在再制造堆焊焊道自动跟踪方面取得突破，为构建磁控电弧自动堆焊理论体系、发展再制造零部件的精确控形堆焊技术提供重要基础。

实现堆焊再制造过程的自动化，所需解决的关键问题之一是自由曲面形状特征的识别。但目前针对再制造堆焊自由曲面的特征识别技术和效果还不尽人意，尤其是对多层多道堆焊和复杂磨损表面堆焊，需要一种低成本的曲面识别方法。.本项目创造性的提出了一种双向磁控电弧式和一种具有磁聚电弧作用的传感器，建立了电弧及熔滴的复合热源模型。获得了熔池形态变化、各节点温度变化及温度梯度对焊缝跟踪及焊接应力变化规律。通过建立的Kriging代理模型，采用多物理场的耦合方法，优化了磁控电弧传感器的结构参数和性能。.提出一种匹配追踪和非参数基函数相的磁控电弧传感特征信号的提取方法，研究了基于变分模态分解和经验小波变换二种埋弧焊信号分析方法。通过对堆焊时温度场的三维有限元动态模拟，建立了磁控电弧传感器空间位姿与复杂曲面特征之间的数学关系。实验结果表明，提取的特征信号更加精确，为堆焊系统的优化提供了新的解决方案。.提出并研究了焊接机器人的焊枪倾角和空间位姿自适应调节、基于谐波分析和向量空间特征分析的焊枪位姿识别等方法。建立了磁控电弧传感器空间位姿与焊接坡口特征之间的数学关系。提出焊枪位姿分级阈值控制策略和姿态协调策略。.研究了堆焊系统主要参数对熔滴过渡和熔池液面形态的影响规律，建立了基于熔池重叠面积三维空间焊缝跟踪精度数学模型，仿真分析表明，焊接时机器人的焊接电流和焊接速度对焊缝跟踪精度的影响显著，对精度模型主要焊接参数优化能够有效的提高焊缝跟踪精度和焊接质量。.通过研究熔滴过渡和焊接熔池液面形态对自由曲面识别的影响规律、研究传感器输出信号与自由曲面特征的相互关系，提出了多层多道堆焊和零部件表面修复的自动化和智能化方法、自寻路径复杂曲面堆焊方法、基于微机械传感器的复杂曲面堆焊方法、堆焊机器人的焊枪空间位姿自适应调节方法等一些实现自由曲面识别以及焊道自动排列的新方法，为其在再制造领域中的应用提供理论依据。