Tool wear states recognition based on integrated neural networks

Cutting tool monitoring is a key technology for automatic, unmanned and adaptive machining. It´s vital to choose right monitoring and recognition methods. Cutting force and vibration are good manners for tool wear monitoring. This paper puts forward techniques of applying frequency band energy decomposition using wavelet packets to extract signal features. And aiming at shortcomings of using single artificial neural network to integrate multi-sensor information, the integrated neural networks based tool wear recognizing process is proposed to accomplish decision-making level data fusion. Experimental results have shown that tool wear diagnostic rate can then be greatly improved.