Research On Condition Monitoring Technology Of Automobile Accessory Drill Wear And The State Of Chip Removal For Deep-hole Drilling Based On DSP

There are many problems in the deep hole processing of Automobile Accessory. In cutting process, due to the presence of chip, lubrication difficulties and problems of the vibration of the tool, it will seriously affect the quality of processing. BTA deep hole drilling is the research object in this thesis, the generation mechanism of tool grinding meal,plugging dust and influence rule of drill pipe torsional vibration,the oil circuit system pressure are analysed. Condition monitoring scheme is established based on drill pipe vibration signal,oil pressure signal, system based on DSP is set up, and the drill pipe vibration signal and the oil pressure signal are successfully acquired?Ddrill pipe vibrations signal and oil pressure signal are comprehensively analysed in domain, frequency domain, time-frequency domain. The change rule of monitoring signal characteristics are studied in whole process of drilling. The energy Q IMF1, the mean of IMF2 and RMS of IMF3 by HHT transform of drill pipe vibration signal are extracted as tool wear state recognition eigenvector of sample, which closely relate to drill wear condition. oil rate, energy Q and peak factor C by time domain waveform of the oil pressure signal are extracted as clearance status identification eigenvector of sample.According to randomness and uncertainty between drilling condition and characteristics vectors, Least Squares Support Vector Machine(LSSVM) of small sample and nonlinear classification is proposed to construct state recognition system, and fluence of penalty factor C and kernel function width parameters ? of LSSVM is studied. The Adaptive Genetic Algorithm(AGA) is employed to optimize penalty factor C and kernel function width parameters ? of LSSVM. condition recognition system of the least squares support vector machine based on the adaptive genetic algorithm(AGA-LSSVM) is constructed.Deep-hole drilling condition monitoring signal data acquisition system is developed based on LABVIEW platform and the upper computer program of reception. display and storage is developed. DSP embedded real-time monitoring system also is developed CCS3.3 environment and software of system initialization, signal processing, condition recognition and warning output is completed. Compared monitoring results with off-line results show the system has high monitoring accuracy and better reliability real-time, and it completely meet the BTA deep hole drilling state of real-time monitoring.