Economic-Statistical Design of an Integrated Triple-Component Model Under Various Autocorrelated Processes

It has recently been proven that integrating statistical process control (SPC), maintenance policy (MP), and production could bring benefits for the entire production system. In the literature of integrated triple-component models, independent observations have generally been studied. The existence of correlated structures in practice put the traditional control charts in trouble. The mixed EWMA-CUSUM (MEC) chart has been developed as an effective tool of SPC for monitoring only the autoregressive (AR) processes. Nevertheless, it has not been extended for moving average (MA) and ARMA processes. Besides, MEC has been designed only based on statistical measures. However, in an imperfect production system, the decision variables of MEC together with the other components should be determined according to the resulting costs and satisfaction of some criteria. This paper proposes an integrated triple-component model by applying the MEC chart for monitoring various autocorrelated processes. Due to the complexity of the model, a particle swarm optimization (PSO) algorithm is employed to reach optimal solutions. The applicability of the model is investigated via an industrial example. The effects of model parameters on the solutions are studied through a sensitivity analysis. Moreover, extensive comparisons and a real data set are provided for more investigations.