大切削力扰动激励作用下机器人制孔系统钻削毛刺形成机理及预测方法

The fixture hole drilling for vehicle frame is the key process in anti-aircraft assembling. The quality of drilling directly affects the assembly quality. Because of the large hole span, drilling process still rely on artificial method which induced low efficiency and poor quality. Because of its good flexibility and high space accessibility, the multi-joint robot drilling system has become an effective way to replace the artificial drilling process for anti-aircraft vehicle frame’s fixture holes. However, the stiffness of multi-joint robot drilling system is 1-2 orders of magnitude lower than machine tool. When drilling a large diameter fixture hole on a stainless steel connection board, the system prone to deformation and chatter by the role of large cutting force, resulting in serious drilling burr. In this project, the influence of disturbance of the dynamic load resulted with the large drilling force on the robot system was studied. First, the mechanism of the burr formation in the drilling process of the weak rigid system under large cutting force disturbance is revealed. The physical model of burr formation is then build. Second, the cyber-physical model for robot drilling system is build according to the physical prediction model. A burr prediction model under the condition of weak rigid dynamic disturbances of anti-aircraft vehicle frame was established based on Sparse Autoencoder Deep Neural Network which is to investigate the key factor to burr formation. Drilling experiments were conducted to verify the theory and model proposed by this project. The research of this project has important theoretical significance and application value for enriching the theory of robot machining system and improving the processing quality and efficiency of the robot drilling system.

地面防空战车车架安装孔钻削是战车装配的关键工序，制孔质量直接影响装配质量。由于安装孔间跨度大，位置分散，目前仍依靠人工制孔，存在缺陷多，效率低等问题。多关节机器人制孔系统因其柔性好、空间可达性高，是代替手动制孔的有效途径。然而多关节机器人制孔系统刚度弱，在奥氏体不锈钢车架连接板上钻削大直径安装孔时，受大切削力作用系统易发生变形和颤振，结果导致钻削毛刺严重。本项目通过研究大钻削力动态激励对机器人制孔系统负载的扰动规律，揭示大切削力激励扰动下弱刚性系统制孔过程毛刺形成机理，建立毛刺生成物理模型;在此基础上,构建基于物理模型的机器人制孔系统信息物理映射模型，采用深度神经网络算法建立毛刺智能预测模型，以辨识复杂动态环境下影响毛刺生成的敏感因素。通过实验对所提出的理论及模型进行验证。本项目的研究为实现机器人制孔毛刺主动控制，提高机器人制孔系统的加工质量与效率，具有重要的理论意义和应用价值。

地面防空战车车架安装孔钻削是战车装配的关键工序，制孔质量直接影响装配质量。由于安装孔间跨度大，位置分散，目前仍依靠人工制孔，存在缺陷多，效率低等问题。多关节机器人制孔系统因其柔性好、空间可达性高，是代替手动制孔的有效途径。然而多关节机器人制孔系统刚度弱，在奥氏体不锈钢车架连接板上钻削大直径安装孔时，受大切削力作用系统易发生变形和颤振，结果导致钻削毛刺严重。本项目通过研究大钻削力动态激励对机器人制孔系统负载的扰动规律，揭示大切削力激励扰动下弱刚性系统制孔过程毛刺形成机理，建立毛刺生成物理模型;在此基础上,构建基于物理模型的机器人制孔系统信息物理映射模型，采用深度神经网络算法建立毛刺智能预测模型，以辨识复杂动态环境下影响毛刺生成的敏感因素。通过实验对所提出的理论及模型进行验证。本项目的研究为实现机器人制孔毛刺主动控制，提高机器人制孔系统的加工质量与效率，具有重要的理论意义和应用价值。项目执行期间已发表期刊论文4篇，申请发明专利1项。多次参加智能制造相关学术会议，并做报告2次。依托本项目培养硕士研究生和博士研究生若干。项目总体执行情况良好，完成了预期目标。