Polyakova M. Segmentation of random processes realizations in automatized systems of technical diagnostics of cutting tools

Automatized systems of technical diagnostics of cutting tools. To increase the productivity of automatized systems of technical diagnostics of cutting tools by means of increasing diagnostics accuracy and automatization level of these systems at the expence of increasing noisestability and threshold selection automatization of procedures of segmentation and identification of random prosesses realizations in automatized systems of technical diagnostics of cutting tools. Methods of research are statistical analysis of random processes and wavelet analysis. Mathematical model of profilograms of cutting edges and vibroacoustic signals of cutting tools as random processes realizatons with segment stationary garmonic differences is proposed. Method of joint detection of changes in mean and changes in garmonic components of random process in wavelet transform domain is improved. This permits to automatize the threshold selection procedure. The method of identification of random processes realizations based on moments of their characteristics changes is proposed. This provides trend and scale invariant. Sphere of use is flexible manufacturing systems.