Heterogeneous recurrence monitoring of dynamic transients in ultraprecision machining processes

In situ monitoring and control of process variations are important for quality assurance in ultraprecision machining (UPM) processes. Recent advancements in sensing and communication technology have fueled increasing interests to develop sensor-based monitoring approaches for anomaly detection in the UPM process. However, conventional approaches are limited in their ability to address the complex dynamics hidden in the nonlinear and nonstationary processes. As a result, it is difficult for them to effectively capture the process variations of UPM. This paper presents a new heterogeneous recurrence monitoring approach to detect dynamic transients in UPM processes. First, a high-dimensional state space is reconstructed from in situ sensing signals. A Dirichlet process (DP) driven clustering approach is then developed to automatically segment the state space into local recurrence regions. Furthermore, a fractal representation is designed to characterize state transitions among recurrence regions and extract novel measures to quantify heterogeneous recurrence patterns. Finally, we integrate a multivariate control chart with heterogeneous recurrence features for in situ monitoring and predictive control of the UPM process. Experimental results showed that the proposed approach effectively detects transitions with a small magnitude, i.e., p = 28 to p = 27 in the Lorenz system, and identifies the shift from stable cutting (R-a = 35 nm) to unstable cutting (R-a = 82 nm) in UPM processes with an average run length of 1.0. This paper presents a novel data-driven DP clustering approach to characterize heterogeneous recurrence variations and link with the quality of surface finishes in UPM processes. This new DP recurrence approach circumvents the need to empirically define local recurrence regions and is shown to have strong potentials for manufacturing process monitoring and control that will increase the surface integrity and reduce rework rates. (C) 2016 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.