An Integrated EPQ Model Of Imperfect Production Process Based On Design Of Control Chart

Traditional EOQ/EPQ models become inadequate and inapplicable in a rapidly developing manufacturing world. As a result, new EOQ/EPQ models that are well adapted to a real manufacturing environment are developed by researchers recently when more practical operation manufacturing issues are taken into account. Traditional EPQ models determines lot size merely considering ordering cost and storage cost generated by a perfect production process, which is a simple assumption far away from real manufacturing environment. And the results given by these models can't provide scientific support to decision-makers. So this paper proposes expanded EPQ models that could satisfy manufacturing requirements and help to make manufacturing decisions based on previous researches.According to real manufacturing conditions, production process in this paper is defined as an imperfect process that involves two production states: In-control state and out-of-control state. Taguchi loss function is applied to estimate quality loss costs of both states. As an effective and widely used SPC tool, control chart is used to monitor and control the production process. Moreover, the concept of maintenance management is also integrated into this model, with SPC tools, delivering good performance on the production process management.This paper develops a three-stage model, which integrates quality loss, control chart and process shift of SPC, and maintenance management step by step. Some detailed information is provided as follows:(1) In the first stage model, process shift is taken as a constant, with control chart and Taguchi quality loss function to calculate the quality loss cost. And the quality loss cost is added into the traditional cost structure consisted of ordering cost and storage cost.(2) Second stage model considers process shift as a variable that follows Raleigh distribution. This assumption is much closer to a real situation. Moreover, control chart with warning lines is used to monitor the production process, which is expected to realize another out-of-control state.(3) In the final stage, an integrated model that takes maintenance management concept is developed based on previous two stage models. This model recommends different maintenance strategies should be implemented to corresponding process state. The economic mode based on ordering cost, storage cost, quality loss cost, inspection cost and maintenance cost makes the cost structure more comprehensive.The goal of the proposed model is to minimize the expected production cost covering five kinds of costs: ordering cost, storage cost, quality loss cost, inspection cost and maintenance cost, though determining optimal decision variables: coefficient of control chart, interval of inspection, sample size, the time of production cycle and the coefficient of maintenance management.Algorithm of Pattern research is applied to solve the model, which is realized by Tool Box of MATLAB 7.0. Conclusion about results is provided after presenting a case example. More discussions are also conducted to compare the three stage models. Compared with the previous researches in this filed, this paper makes following contributions:(1) Define an imperfect production process, which covers the gap between traditional models and real manufacturing situations.(2) Assuming process shift occurs to the production process, this paper discusses what impacts have on the model when this process shift is considered as a constant or a variable that follows Raleigh distribution.(3) Take SPC concept into account by establishing link between control chart and Taguchi quality loss function. Economic design of control is investigated without compromising its statistical design criterions.(4) Integrating SPC concept and maintenance management idea into the model, the paper will provides right strategies depending on what conditions the manufacturing environment is actually under.(5) This paper determines decision variables that are mostly related to manufacturing actions. So the results will give lots of effective support to decision-makers when they make manufacturing schedules and plans.