Tool Wear Condition Monition Based On AE And Vibration

Tool is the executor of the metal cutting. So Tool wearing is an inevitably phenomenon occurring in the process of metal cutting. The transformation of tool-state directly causes product quality declining, production cost increasing, and gradually affect the capacity of market competition of products. Tool condition online monitoring has a great meaning to improving product quality, decline product cost and improving productivity. In order to solving this problem, this paper's research as follows:This paper, firstly, established the tool condition monitoring signal sampling system based on acoustic emission sensor and vibration sensor. After experiments, we got a lot of signals in different tool wearing conditions. Secondly, we analyze these signals in time domain, frequency domain as well as wavelet packet transformation. Then we used correlation coefficient method to select the characteristic quantities which have a close relationship with tool wearing value. Thirdly, this paper analyzed the features of BP neural network which is applied to tool condition monitoring. And a new method, three times spline weight-function neural network, is first used into tool condition monitoring. This method improves the recognition speed and multiplication of monitoring system, therefore it is suitable for on-line monitoring. At last, this paper aimed at the problem of poor reliability of monitoring system. So it used integrated neural network to build the allusive relationship between tool wear and signal features.The researches on the experiment design, signal collection, signal analysis, feature extraction and feature selection and pattern recognition are exploratory in this thesis. The precision and reliability of monitoring system are improved.