An EWMA Control Chart for Multivariate Functional Data
摘要
This paper introduces a novel functional control chart that extends the classical exponentially weighted moving average (EWMA) control chart from univariate scalar data to multivariate functional data. Through extensive Monte Carlo simulations, the proposed control chart demonstrates superior performance compared to competing monitoring schemes. Its practical utility is illustrated through a case study in automotive manufacturing, where it is employed to monitor the quality of resistance spot welding processes by analyzing dynamic resistance curves.