An intelligent pattern recognition algorithm and its application in cutting tool condition monitoring process

摘要

An intelligent tool wear monitoring system for metal cutting process will be introduced in this paper. The system is equipped with four kinds of sensors, signal transforming and collecting apparatus and a micro computer. A knowledge basedintelligent pattern recognition algorithm has been developed. The fuzzy driven neural network can carry out the integration and fusion of multi-sensor information. The weighted approaching degree can measure the difference of signal features accuratelyand ANNs successfully recognize the tool wear states. The algorithm has strong learning and noise suppression ability. This leads to successful tool wear classification under a range of machining conditions.