Grinding chatter detection method based on stack type self-encoder and support vector machine

Abstract

The invention discloses a grinding chatter detection method based on a stack type self-encoder and a support vector machine. Firstly, tag data and test data are used for jointly training the stack type self-encoder, then, the trained stack type self-encoder is used for carrying out processing on the tag data to obtain tag characteristics which are used for training the support vector machine; finally the trained stack self-encoder takes test data characteristics extracted through the test data as the input of the trained support vector machine, fault diagnosis is carried out, and a diagnosis result is obtained. The method has generality, the chatter judgment is not affected by machining equipment and machining parameters, the method is suitable for grinding chatter detection, and can be used and popularized in a machine tool enterprise, the intelligent machine tool development needs are met, and the wide application prospect is achieved.

Description

technical field [0001] The invention relates to the technical field of machine tool chatter detection, in particular to a grinding chatter detection method based on a stacked autoencoder and a support vector machine. Background technique [0002] Grinding, as a common final processing technology, has a decisive impact on product processing efficiency, geometric accuracy and surface quality. The occurrence of grinding chatter will aggravate the wear of the grinding wheel, generate abnormal noise, reduce the life of machine tool components, and reduce the processing accuracy and surface quality of the product. Quality, the evolution of grinding chatter behavior is staged, usually including three stages of early stage, rising period and stable development. In the grinding process, chatter chatter can be avoided only if chatter chatter is detected in the early stage and suppression measures are taken. It will damage the processing accuracy and surface quality of the product. At ...

AI Technical Summary

Problems solved by technology

[0002] Grinding, as a common final processing technology, has a decisive impact on product processing efficiency, geometric accuracy and surface quality. The occurrence of grinding chatter will aggravate the wear of the grinding wheel, generate abnormal noise, reduce the life of machine tool components, and reduce the processing accuracy and surface quality of the product. Quality, the evolution of grinding chatter behavior is staged, usually including three stages of early stage, rising period and stable development. In the grinding process, chatter chatter can be avoided only if chatter chatter is detected in the early stage and suppression measures are taken. It will damage the processing accuracy and surface quality of the product. At present, the existing online detection methods for grinding chatter have poor versatility, which is mainly reflected in the sensitivity of the chatter threshold to processing equipment and processing parameters. Therefore, it is necessary to have a detection method that can extract chatter Characterization of Vibration-Independent Machining Parameters