Research On Drill Wear Monitoring Technology Based On HMM By Using The Force And Noise Signal

With the development of automatization manufacturing equipment such as MC and FMS etc, cutting tool condition monitoring becomes a key technology to realize automatization in machining process. So it is significant to study tool wear and breakage condition monitoring techniques. With the drilling process as the research objective, the feature extraction of drilling force and noise signals during the drill wear monitoring as well as the drill wear condition recognition was studied theoretically in this article; the drill wear monitoring experimental system based on HMM with the drilling force and noise as the monitoring signals was established too.The approaches for non-steady signal processing based on STFT, wavelet transform, wavelet packet transform were discussed; the relation between the force, noise signal and condition during drilling process was researched. It had been found that the power spectrums of drilling force and noise signal showed good correlation with drill wear. According to the theory of wavelet packet power spectrum, wavelet packet power spectrum of force and noise signals had been obtained successfully and the features monitoring drilling condition had been extracted triumphantly.From the point of fractal, the internal relationship between wavelet transform and the fractal theory was researched; an approach of defining fractal dimension which took the wavelet as a standard was put forward.; A kind of wavelet fractal dimension which uses the absolute plus of the signal as measure parameters was established; the relation between the change law of the fractal dimension and the tool wear was discussed through the research on the cutting force of the whole process and the wavelet fractal dimension of noise signals. The result shows that the substance of the signal could be described effectively by means of wavelet fractal dimension; and there is a good correlation between the change of the signal's fractal dimension and the wearing of the drill. As a result, the monitoring of the drill tool condition can be accomplished successfully by the means proposed above.The theory of HMM had been introduced into the field of the drill wearing condition monitoring. Many factors on HMM (including the structure, the initial of the parameters and the Vector Quantization technique) had been studied particularly. The HMM drill wear monitoring system based on the wavelet packet power spectrum had been established later and the monitoring effect of the HMM drill wear monitoring system had been tested by many experiments. The experimental result indicates that there is a strong relation between the feature of the signal and the drill wearing condition; the trend of the drill wear condition could be reflected effectively by the monitoring system based on HMM; and the monitoring and recognition about the drill wearing...