基于分形维数的刀具状态在线监测新方法  

New Method of On-Line Monitoring Tool Wear States Based on Fractal Dimension

作　　者：王忠民[1] 王信义[1] 陈爱弟 杨大勇[1] 贾玉平[1] 

机构地区：[1]北京理工大学机械工程与自动化学院,北京100081

出　　处：《北京理工大学学报》2000年第4期441-444,共4页Transactions of Beijing Institute of Technology

基　　金：部级基金

摘　　要：以分形几何理论为基础 ,对刀具不同磨损阶段声发射信号的分形特征进行分析 .提出了计算非完全分形体信号波形的关联维数时尺度范围的确定方法 ,分析了声发射信号在刀具磨损过程中分形维数的变化特性 .刀具磨损切削实验数据表明 ,声发射信号的分形维数受切削参数变化影响较小 ;分形维数反映了声发射信号的几何特征 ,其大小能较好地反映刀具的不同磨损状态 .实验结果表明 ,该方法能正确地实时在线监测刀具的不同磨损状态 .On the basis of the theory of fractal geometry, the fractal dimension characters of AE signal at different tool wear states were analyzed, an algorithm was provided to decide the scale range which is necessary to calculate the dimension of non fully fractal form. The experiments showed that the fractal dimension of AE signal is slightly effected by the changes of cutting parameters; The variance of the fractal dimensions, which reflects the geometric characters of AE signals, has the same tendency as that of the flank wear, The monitoring system, which takes the fractal dimension of the AE signal as the feature, can identify the different tool wear states more correctly under different cutting conditions.

关 键 词：分形维数 声发射 刀具磨损 在线监测 刀具状态 

分 类 号：TG502.3[金属学及工艺—金属切削加工及机床]