Numerical control machine tool cutter wear loss online prediction method based on spindle currents and vibration signals

Abstract

The invention discloses a numerical control machine tool cutter wear loss online prediction method based on spindle currents and vibration signals. The method is characterized in that multiple cuttersof the same model are utilized to repeatedly carry out trial operation machining under the same working condition on a numerical control machine tool spindle provided with a sensor, and original machining wear data are measured; according to three wear stages of the initial rapid wear stage, the normal wear stage and the rapid wear stage of the cutters, an optimal feature set for training is extracted from each wear stage, and a support vector regression machine prediction model is trained; and finally the trained model is adopted for online real-time prediction of cutter wear loss in the actual machining process. According to the method, through the data obtained by repeated experiments through the multiple of cutters of the same model under the same condition, characteristic parameters,related to cutter wear, in the wear stages in original signals can be fully mined, the relational degree between the characteristics and the cutter wear loss is enhanced through a post-processing mode, and therefore, the constructed support vector regression machine prediction model can obtain high prediction precision and good generalization performance.

Description

technical field [0001] The invention relates to a technology in the field of mechanical processing, in particular to a high-precision online prediction method for tool wear of a numerically controlled machine tool based on spindle current and vibration signals. Background technique [0002] CNC machine tools are important basic equipment of modern advanced manufacturing technology, and the downtime of machine tools caused by tool failure accounts for about 1 / 5-1 / 3 of the total downtime of machine tools. For a CNC machine tool equipped with a relatively mature tool monitoring system, its downtime can be reduced by 75%, and the processing efficiency can be increased by 10-60%. The existing tool condition monitoring technology can generally achieve more accurate prediction of the tool wear status category, but the prediction accuracy of the wear amount is not high, and the real-time performance and popularization are poor, so it is difficult to apply it to industrial production...

AI Technical Summary

Problems solved by technology

[0003] The present invention aims at the defects of low prediction model accuracy of the existing tool monitoring technology and poor generalization of the model established based on the experimental data of a single tool, and proposes an online prediction method for tool wear of CNC machine tools based on spindle current and vibration signals. The data obtained from repeated experimen

Images