Control chart forecasting: A hybrid model using recurrent neural network, design of experiments and regression

Abstract

Recurrent neural network (RNN) is an efficient tool not only for modeling production control process but also for modeling services. In this paper the combination model of RNN, regression and stepwise regression analysis (SRA) were employed in order to predict the variables of process control chart. Therefore, one maintenance process in workshop of Esfahan Oil Refining Co. (EORC) was taken for illustration of hybrid model. First, the most important factors on forecasting response time as inputs were selected according to SRA. Then, the regression was made for predicting the response time of process based upon obtained inputs, and then the error between actual and predicted response time as output along with input were used in RNN. Finally, according to predicted data from combined model, it is scrutinized for test values in statistical process control whether forecasting efficiency is acceptable. Meanwhile, design of experiments (DOE) was set so as to optimize the RNN in training process of it.