飞机结构件内腔机器人自动打磨工艺

Technology of Robotic Automatic Grinding Internal Surface in Aircraft Structural Parts

飞机结构件的内腔结构复杂,打磨工艺难以控制,为提高打磨表面质量和打磨效率,设计了一种具备轴向/径向恒力控制的末端执行器,并搭建了自动打磨实验台。运用正交实验对磨具粒度、打磨压力、进给速度等打磨工艺参数进行了极差分析和方差分析,得到了显著因素和最优水平组合。以表面粗糙度和去除深度为评价指标,建立BP神经网络预测模型,并对训练后的模型进行实验验证。实验获得的实测值与预测值的相对误差在5.5%以内,表明该模型对表面粗糙度和去除深度有良好的预测能力,可有效地提高打磨工艺设计的效率。

As complex structure existed inside aircraft structural parts,the process of grinding is difficult to control.In order to improve the surface quality and efficiency of grinding,an end-effector with axial/radial constant-force control function was designed and an experimental platform of automatic grinding was built.Range and variance analysis were conducted on the process data of abrasive granularity,grinding pressure and robotic moving speed by means of orthogonal experiment,and the significant factors and optimum technological conditions for the best quality were determined.Taking the surface roughness and grinding depth as the evaluation index,a back propagation neutral network prediction model was built,trained and verified by experiment.The relative error between the prediction value and the actual value got in the verification experiment was within 5.5%,which shows that this model has good predictive ability for surface roughness and grinding depth and can effectively improve the efficiency of grinding process design.