A closed-loop multivariable quality control scheme for injection molding process

Closed-loop quality control by adjusting multiple process parameters can enlarge the operating space; therefore, it can deal with more severe quality deviations in the quality control system. A closed-loop multivariable quality control scheme is presented for injection molding process. The initial process settings are determined by DOE based parameter optimization. The multilayer group method of data handling (GMDH) algorithm is used to develop the quality model. The sequential quadratic programming (SQP) method is adopted in the quality controller design, where the optimal parameters are calculated cycle by cycle according to the observed quality deviation. Quality measurements are provided by a multiway partial least square (MPLS) regression model. The illustrative example can show the feasibility of the proposed quality control scheme.