基于切削力和切削温度实施刀具实时监测的研究-机械制造及其自动化专业论文.docxVIP

优秀毕业论文 精品参考文献资料 青岛理工大学工学硕士学位论文Abstract 青岛理工大学工学硕士学位论文 Abstract As the key technique in automatic and unmanned machining processing，the in—process tool condition monitoring is a very active research field both at home and abroad．The tool wear is severely affecting the product quality and efficiency． Make the flank wear as the monitoring object．Realize the in-process tool condition monitoring using sensor fusion of cutting force and cutting temperature．The factors which are concerned谢th cutting force and cutting temperature signals are confirmed． The orthogonal test which is based Oil neural network and signal factor test are designed． The cutting force and cutting temperature signals has been analysed in frequency domain and time domain and wavelet analysis．The reasons are analysed how feature changes are caused by changing parameters．It is very apparent that three cutting factors affect the cutting force and cutting temperature signals，from the research of influence factors on cutting force and cutting temperature．The free-machining metals have little influence on cutting force features．But for low machinability metal，the effects are great． The cutting force features rapidly change．The influence of workpiece materials on cutting temperature is very apparent． The features are found which has a significant correlation witll flank wealfrom the single·factor test of flank wear．They are mean value and wavelet packet energy spectrum of cutting force，mean value and frequency spectrum of cutting temperature．They Can easily reflect the changes of flank wear． Using BP neural network training and the data getting from orthogonal test method， the mapping relation between cutting force，cutting temperature features and flank wear． The network is corrected by the data getting from random experiment．The training error is le一005．The training precision is 1 00％．It ensures that the network has high-precision． This study solves a series of problems of tool wear monitoring in experimen