A0357
Title: A semi-parametric CUSUM scheme using copula formulation for higher-order autocorrelated processes
Authors: Yang-Li Liao - Feng Chia University (Taiwan) [presenting]
Abstract: In the Industry 4.0 era, rapid operations, characterized by the four Vs (volume, velocity, variety, and veracity), often encounter challenges related to autocorrelation in data streams. This manuscript presents a novel monitoring approach designed for autocorrelated processes, overcoming constraints on the order and magnitude of autocorrelations and distributional assumptions. The proposed phase II semi-parametric copula-based cumulative sum (CUSUM) control chart effectively detects shifts in location within continuous autocorrelated processes without relying on distributional assumptions. Average run length (ARL) performances against a distribution-free strategy demonstrate the proposed method's efficacy, showcasing adaptability in handling various autocorrelation magnitudes. Application of this CUSUM method in the selective laser melting (SLM) process produces prompt signals, validating its effectiveness in real-world datasets.
