Tool running state reliability evaluation method based on EMD

Abstract

The invention provides a tool running state reliability evaluation method based on EMD. The method includes the steps that wavelet denoising is conducted on state signals such as a force signal, a vibration signal and an acoustic emission signal which are collected in the tool machining process, then after EMD is conducted, IMF components related to tool abrasion are extracted, the mean value of amplitudes of the IMF components is used as an information index of the tool running state, a characteristic matrix is built, a sub-space of the characteristic matrix at the normal state and a sub-space of the characteristic matrix at the current state are constructed through a sub-space construction method based on kernel principal component analysis, and a main included angle of the sub-spaces is calculated and used as the reliability of a tool. The method is not based on a large sample condition, is completely suitable for small sample tools with failure data, is especially suitable for evaluating the reliability of tools which are high in manufacturing cost and are made specially, and is accurate in evaluation result.

Description

technical field [0001] The present invention relates to the related fields of signal processing technology and fault diagnosis, specifically a method based on EMD decomposition (Empirical Mode Decomposition) by processing vibration signals, acoustic emission signals and force signals in the operating state of the tool. A method for evaluating the reliability of cutting tool operation status. Background technique [0002] CNC machine tools are the mainstream equipment in the contemporary machinery manufacturing industry. Our country can develop CNC machine tools with technical parameters consistent with the international level, but in terms of precision stability and reliability of machine tools, it is far from the advanced level of foreign countries. The average time between failures is almost The difference is 1 times. Improving the reliability of my country's CNC machine tools has become an important guarantee and an urgent problem to be solved for improving my country's eq...

AI Technical Summary

Problems solved by technology

Aiming at the problem that expensive tools under single working conditions are difficult to evaluate with traditional reliability methods, the present invention proposes a reliability evaluation method based on EMD decomposition of the operating state of the tool, according to the similarity between the current state and its normal state The reliability index is obtained without relying on historical failure data, and the real-time evaluation of the reliability of the tool operating state under a single working condition can be realized. The core process is to obtain the wear characteristics of the state information during the tool operation and establish the state subspace ;Calculate the main included angle between the normal state and the current state subspace base vector, the main included angle reflects the degree to which the tool performance deviates from the normal state at the current moment, and reflects the ability of the tool to complete the normal specified functions; finally, the reliability is defined according to the main included angle Indicators for reliability evaluation of tool operating status

Images