Industrial Applications of Online Machining Process Monitoring System

An online machining process monitoring system has been constructed by taking advantage of the new achievements in data acquisition, sensor technology, and signal processing. Firstly, the architecture of monitoring system with the capability of automatic online acquisition, presentation, and analysis sensory signals is designed. Secondly, wavelet transform is further explored to decompose sensory signals into static and dynamic components for the purpose of extracting distinctive features associated with different tool malfunctions. Thirdly, by conjunction with the wavelet transform, univariate and multivariate statistical process monitoring techniques are proposed to construct the thresholds of malfunction-free machining zone. Short time Fourier transform is further introduced to detect the onset of chattering in machining processes. Finally, the effectiveness of the developed techniques and monitoring system has been demonstrated by experimental results obtained from the extensive industrial machining trials.