机器人高速双丝弧焊多源信息融合与智能参数匹配策略研究

Multi-parameter input, as well as high-speed time-variation in real time are two significant characteristics during robot manipulation based double-wire arc welding process. Further researching the mechanism of high-speed double-wire arc welding and corresponding technic rule is crucial for achieving the reasonable parameters matching and then improving the quality of robot welding product. Our group will establish a welding quality prediction model based on the characteristics of double-wire arc welding for robot manipulation, through theoretical modeling, simulation, experimental verification and intelligent optimization, and then explore the unified intelligent parameters matching control strategy. This research will be composed of the following aspects:.1.The credible quality criterion for welding quality based on electrical and electrical arc sound signals will be confirmed, and then a corresponding multi-resource information merging evaluation model will be established based on support vector machine. Then an approximate welding quality evaluation system can be designed based on expert experiences and national standards, and the error control can be achieved, as well as the model can be optimized based on the system. .2.The effects of different robot moving parameters, technic conditions and welding parameters on the input energy and welding quality will be explored, and then establish a “technic rule-input energy-welding quality” intelligent prediction model. .3.The unified parameters automatic matching control strategy will be explored based on average welding current, and the parameter-matching model will be optimized. Then the goal of realizing the automatic setting of technic parameters can be achieved..The achievement of this research can serve taking knowledge of the mechanism of robot manipulation based double-wire arc welding and improving the product quality of robot welding, and it can also supply the theoretical foundation and technical support for improving the current welding technology based on robot manipulation.

机器人双丝弧焊具有多输入、高速时变的特点，深入研究机器人高速双丝弧焊工艺机理，实现合理参数匹配是提高机器人焊接质量的难点与关键问题。课题组拟通过理论建模、仿真模拟、实验验证、智能优化的方法建立具有机器人高速双丝弧焊特点的焊接质量预测模型，并探索一元化机器人高速双丝弧焊的参数智能匹配策略。具体研究内容为：.1.设计基于电信号、电弧声的焊接质量定量评价指标，建立基于支持向量机的多源信息融合评价模型；结合专家经验、国家标准设计焊接质量模糊评价系统，并据此进行误差控制，优化评价模型。.2.研究不同机器人运动参数、工艺条件、焊接参数对输入能量、焊接质量的影响，建立“工艺规范-输入能量-焊接质量”智能预测模型。.3.研究基于平均电流的一元化参数智能匹配控制策略，优化参数匹配模型，实现焊接参数智能匹配。.研究成果对了解机器人双丝弧焊机理，提高焊接质量，推动机器人焊接智能化技术发展提供理论依据与技术支持。

机器人双丝弧焊具有多输入、高速时变的特点，深入研究机器人焊接规律以及机器人高速双丝弧焊工艺机理，实现合理参数匹配是提高机器人焊接质量的难点与关键问题。项目组综合使用理论建模、仿真模拟、实验验证、智能优化等各种方法建立具有机器弧焊特点的焊接质量预测模型和方法，进行了大量的规律实验探索一元化机器人弧焊的参数匹配策略。具体研究内容为：.1.研究基于电信号、熔池图像、焊缝的焊接质量评价指标，建立模糊评价、灰度理论等方法的焊接质量评价模型；结合专家经验、国家标准设计焊接质量模糊评价系统，并据此进行误差控制，优化评价模型。.2.研究不同机器人倾角、运动参数、工艺方法、焊接参数对输入能量、焊接质量的影响，建立“工艺规范-焊接质量”对应关系，指导和优化焊接工艺。.3.研究探索了机器人电弧增材焊接等先进焊接技术特点与焊接规律，指导了电弧增材焊接的工艺参数选用。.4.研究基于平均电流的一元化参数智能匹配控制策略，优化参数匹配模型，实现焊接参数智能匹配。.研究成果对了解机器人弧焊机理，掌握机器人焊接特点和规律，提高焊接质量，推动机器人焊接技术发展提供理论依据与技术支持。