Study On Lathe Noise Resource Identification And Design Of Analysis Software

The analysis of lathe noise signal can not only identify noise source, but also diagnose lathe fault. At present, most lathe manufacturer in our country analyzes the lathe working status by worker's ear, but the efficiency of this method is very low. The lathe noise frequency is very complex and the analysis methods provided by the instrument are limited. So it often can not get much more information of the lathe. Around the lathe noise resource identification and analysis software design this thesis takes the following works:First of all, on the basis of summing up the general lathe noise research at home and abroad, the thesis constructs the lathe noise signal analysis system. On the demand of the system,the microphone and sound card are chose as system hardware. The LabVIEW8.2 Chinese edition is chose as software development platform after comparing the advantages and disadvantages of VC + +, MATLAB and LabVIEW.Secondly, the thesis analyzes the major factors that influence the lathe noise and the lathe noise signal characteristics. It describes the lathe noise signal processing methods from time domain, frequency domain and time-frequency domain and compares their advantages and disadvantages. To the lathe noise signal characteristics, it selects the proper lathe noise signal processing methods from time domain, frequency domain, time-frequency domain. It focuses on the study of lathe noise resource identification method which uses wavelet packets and grey theory. The lathe noise signal is decomposed to five layers by using wavelet packets.Then the energy in the first fourth frequency band is extracted to use as feature vector to make grey relational analysis. The lathe noise resource is identified in terms of grey relational degree. Finally, the lathe noise signal analysis software is designed. The lathe noise signal is analyzed comprehensively by the software. The accuracy and reliability of the software is testified. The effectiveness of the lathe noise resource identification method based on wavelet packets and grey theory is proved by experiment. It is a useful attempt for the lathe noise resource identification automatically.