A fuzzy-nets tool-breakage detection system for end-milling operations

This paper describes a new approach, the fuzzy-nets system, for monitoring tool breakage in end-milling operations. The fuzzy-nets tool-breakage detection (FNTBD) system has a self-learning capability to generate rule bases and to fine tune the term sets of each linguistic variable to the appropriate level of granularity. A self-learning algorithm for developing the FNTBD system consists of five steps:

1. Divide the input space into fuzzy regions.
2. Generate fuzzy rules from given data pairs through experimentation.
3. Avoid conflicting rules based on top-down or bottom-up methodologies.
4. Develop a combined fuzzy rule base.
5. Determine a mapping system based on the fuzzy rule base.

Learning is accomplished by fine-tuning the parameters in the fuzzy-nets system within the on-line learning capability. Following establishment of the rule base, the performance of the FNTBD system is examined for an end-milling operation. It was observed and verified experimentally that this new FNTBD approach can successfully detect tool breakage in end-milling operations.