复杂型面精铣加工过程状态监测与早期诊断研究

Complex shaped components have been widely used in the paramount field of the national economy, such as aviation, aerospace, automobile and so on. Due to coupling effects of the weakly rigid system and complexity of the cutting path during finish milling process, tool wear and chatter are easily to appear. Aiming at the finish milling of typical structure and cutting process, this research intends to establish the theories and methods of online evaluation and diagnosis based the mechanism modeling of multi-source sensor signals and fusion feature extraction. The research mainly include: (1) Synchronization sampling and preprocessing of multi source signal based on position information and angular information. (2) Instantaneous milling force modeling and recognition of time varying force coefficients based on measured signal for five-axis milling. (3) Non-stationary parametric modeling and decomposition of acoustic emission signal based on the research of instantaneous cutting thickness. (4) Mode coupling mechanism analysis and the characterization method in thin wall part milling with complex path. Based on the above research, early tool wear and chatter during finish milling process are online recognized based on feature fusion. Therefore, online monitoring and early diagnosis theory for the finish milling process of the complex surface in working condition can be constructed correspondingly.

复杂型面零部件在航空、航天、汽车等国民经济重要领域中有着广泛的应用。在其精铣过程中，由于受到工艺系统弱刚性、时变性以及复杂走刀路径的耦合作用，铣削刀具容易磨损，并容易出现切削颤振。本课题拟针对典型结构与工艺下的精铣加工工序，基于多源传感器信号机理建模、融合特征提取等理论与方法实现复杂型面精铣过程中的在线监测。研究内容包括：研究包含刀具位置和角度信息的多传感器信号同步采集技术及信号预处理方法。研究五轴铣削过程中的瞬时切削力建模及基于实测信号的时变切削力系数辨识理论与方法。研究瞬时切削几何对于声发射信号的影响规律及其非平稳参数化建模与分解。研究薄壁件复杂铣削路径过程中的振动模态耦合规律及其表征方法。在此基础上基于多传感器特征融合实现复杂工艺路径下的铣削刀具磨损与切削颤振的在线辨识与诊断，从而建立一套面向实际加工的复杂薄壁零部件精铣过程在线监测与早期诊断理论与方法。

随着科技的不断发展，复杂型面零部件在航天、航空、汽车等国民经济重要领域的应用越来越广泛。在其精铣过程中，由于受到工艺系统弱刚性、时变性以及复杂走刀路径的耦合作用，铣削刀具容易磨损，并容易出现切削颤振，因此，对于复杂零部件的铣削加工，尤其对最后的精铣工序而言，如何对加工过程的刀具磨损和切削颤振进行早期的辨识与预测对于保证复杂型面零部件的加工质量、提高其使役性能具有重要的研究意义。. 为满足复杂型面零部件铣削过程状态监测与早期诊断的需要，本课题研究内容包括：铣削过程中的切削机理与切削力建模及时变切削力系数辨识研究；基于离散小波分析联合时频分析的声发射信号分解与特征提取；基于多传感器融合和毗邻网格搜索聚类算法的钛合金转角周铣刀具磨损监测与诊断；基于多传感器融合的复杂路径条件下切削颤振的在线辨识与监测。. 基于上述研究内容，共发表高水平论文12篇，申请发明专利6项，申请软件著作权1项；获得省部级科技进步一等奖1项，重要研究结果如下：通过建立复杂工况下的铣削实验，研究了复杂型面铣削过程中多传感器信号随非平稳工况以及加工状态的变化规律。揭示了瞬时切削厚度与刀具接触区域的周期性与非周期性变化对力、振动和声发射等多传感器信号的影响规律。研究了多传感器信号的机理建模与多源信息融合理论，实现了刀具微小磨损与颤振的早期辨识。建立了非平稳工况下的精铣过程加工状态在线监测与早期诊断理论与方法，提高了复杂型面铣削加工的表面质量与加工效率。