Tool wear states recognition based on frequency-band energy analysis and fuzzy clustering

Cutting tool monitoring is a key technology for automatic, unmanned and adaptive machining. It's vital to choose right feature extracting and recognition methods. By using cutting vibration monitoring and diagnostics technique to monitor tool wear states, this paper puts forward techniques of applying frequency-band energy decomposition using wavelet packets to extract signal features of cutting vibration. This method makes the extracted features sensitive to tool wear and the sensitivity to cutting parameters minimum. The recognition method for tool wear sates was studied through fuzzy clustering. Affinities between the known state and unknown state can be obtained through the fuzzy clustering. Then, tool wear states can be recognized by those affinities. Experimental results show that the tool wear monitoring is achieved by using these feature extracting and recognition methods.