Profile monitoring of reflow process using approximations of mixture second-order polynomials

Abstract

Lately, profile monitoring has received considerable attention in the statistical process control research field. This paper proposes a novel monitoring framework for the reflow process data, which uses two goodness-of-fit criteria to select the change points in the mixture polynomial model. Among change points, the mixture second-order polynomials are utilized to piecewisely approximate the nonlinear profile data of the reflow process. The well-known Hotelling T2 and proposed EWMA4 control charts are then employed to monitor the parameter estimates. The experimental results demonstrate that the proposed monitoring framework presents better performances in detecting outlying profiles than the conventional methods in phase I. In phase II, the performance of the proposed framework is assessed in terms of the out-of-control average run length.