基于特征的复杂结构件数控加工刀具状态实时辨识方法

发布时间：2019-03-16 20:05

【摘要】：基于切削力、振动等信号监测是目前数控加工刀具状态实时辨识的有效手段,而数控加工过程中的监测信号不仅跟刀具状态本身相关,还受工件的几何形状、工艺参数等因素的影响,导致数控加工过程中刀具状态辨识准确性差,尤其是对于多品种、小批量生产的复杂结构件,以上问题更为严重。针对以上问题,本文对基于特征的数控加工刀具状态实时辨识方法进行了深入研究。论文主要成果如下:(1)分析了数控加工刀具状态监测的常用传感器,选择了切削力作为本文刀具状态监测的信号,并通过实验方法分析了加工特征对切削力监测信号的影响,进而提出了基于特征的复杂结构件数控加工刀具状态辨识方法,建立基于特征的刀具状态辨识信息模型,分析了基于特征辨识刀具状态的优势。(2)针对数控加工过程中监测信号不仅与刀具状态相关,同时也受工件几何形状和工艺参数影响的问题,分析了加工特征和刀具状态对监测信号影响,建立了加工特征几何形状数控程序监测信号之间的关联关系,实现了刀具状态监测信号与加工特征之间的实时关联,使刀具状态辨识能够依据不同的加工特征进行分类分析。(3)研究了基于特征的监测信号处理方法及刀具状态实时辨识方法。针对不同飞机结构件加工特征提取不同的信号敏感量,结合加工特征几何信息及工艺信息构建刀具状态辨识向量,使用K-Means聚类算法对其分类,实现了基于特征的复杂结构件数控加工刀具状态辨识。(4)基于CATIA/CAA与LABVIEW平台开发了基于特征的复杂结构件数控加工刀具状态实时辨识系统并在实验中得到验证。
[Abstract]:Signal monitoring based on cutting force and vibration is an effective method for real-time identification of cutting tool status in NC machining. The monitoring signal in NC machining process is not only related to the tool state itself, but also affected by the geometric shape of the workpiece. Because of the influence of process parameters and other factors, the accuracy of tool state identification in NC machining process is poor, especially for complex structural components which are multi-product and small batch production, the above problems are more serious. In order to solve the above problems, the real-time identification method of NC machining tool status based on feature is deeply studied in this paper. The main achievements of this paper are as follows: (1) the commonly used sensors of NC machining tool condition monitoring are analyzed, and the cutting force is selected as the signal of tool condition monitoring in this paper. The influence of machining features on cutting force monitoring signal is analyzed by experimental method, and then a feature-based tool state identification method for NC machining of complex structural components is proposed, and a feature-based tool state identification information model is established. The advantage of identifying tool state based on feature is analyzed. (2) the monitoring signal in NC machining is not only related to tool state, but also affected by workpiece geometry and process parameters. The influence of machining feature and tool state on monitoring signal is analyzed. The correlation between machining feature geometric shape monitoring signal and machining feature is established, and the real-time correlation between tool condition monitoring signal and machining feature is realized. The tool state identification can be classified according to different machining features. (3) the feature-based monitoring signal processing method and the tool state real-time identification method are studied. According to the machining features of different aircraft structures, different signal sensitive quantities are extracted, and the tool state identification vector is constructed by combining the geometric information and process information of machining features. The tool state identification vectors are classified by K-Means clustering algorithm. A feature-based tool state identification system for NC machining of complex frame members is developed. (4) A real-time tool state identification system based on CATIA/CAA and LABVIEW for NC machining of complex frame members is developed and verified in experiments.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG659

【参考文献】