机器人点焊多信息融合及控制

点焊质量一直是国内外焊接国界学者致力研究的重要课题之一。目前尚未得到圆满的解决。点焊过程的复杂性、熔核形成过程的不确定性，尤其是点焊过程采用的质量监控方法的单一性决定了点焊质量控制是一项非常困难的任务。本文分析了基于单一传感器信息资源来描述焊点熔核形成过程存在的局限性，首次提出了采用多传感器信息融合控制点焊质量的思想。通过参数选择、特征信息提取，建立了点焊过程多信息融合模型。采用结合法、平均法和决

Multi-sensor Information Fusion and Control of Robot Spot Welding

Spot welding quality is always one of questions under study,to which many welding scholars in the world devote themselves.It has not been solved satisfactorily at present.The complex nature of spot welding process, undeterminedness of grown nugget and singleness of control method decide that stabilizing quality is a much difficult task.The limitations of describing how the nugget growing up in view of single sensor information source is analyzed in the paper.An idea of using multi sensor information fusion for spot welding quality control is presented for the first time.The information fusion model for weld spot nugget is established through parameter selecting and characteristic information extracting. Heat and time adjustment of spot welding process are made through the best fusion arithmetic adapting combined,average and decision making methods.The spot welding experiments indicate that the new quality control method can better compensate various factors influence in the welding process and assure the spot welding quality stabilization.