Queuing Network Modeling And System Performance Analysis Of Flexible Flow Shop With Material Handling Unit

Make-to-order manufacturing business, which is driven by customer orders, is an important part of our manufacturing industry. However, personalization and uncertainties of customer orders make the production process in the enterprises full of stochastic factors. Meanwhile, delivery requirements is more stringent than usual because of the increasingly market competition. So the exact capacity configuration of facilities is difficult. It is not only to consider the fluctuation of orders and the limit of due date, but also to avoid the waste caused by over-investment of facilities. Thus, MTO managers are forced to run the production more effectively with advanced scientific theories. Before the optimization of all kinds of facilities configuration, the abstract production model should be built appropriately. Therefore, the queueing network model on flexible flow shop with material handling unit is studied in this paper which is against the background of a deluxe package manufacturer, a MTO enterprise with single or small quantities of orders.Material handling unit in flexible flow shop can be abstracted as a batch process whose lot size is a random variable. As the batches in unit are not identical, traditional queuing network model and its approximate solution method cannot be used for these problem directly. So the state space decomposition method has been improved to solve this model of flexible flow shop with material handling unit approximately, according to the principle of stochastic process. The main works of this paper are following:Firstly, a mathematical model which can describe the performance of system, based on the common features of the flexible flow shop with material handling unit has been built up. The original problem is divided into the problem that the unit that has only one type of transporters as well as the problem that the unit has several types of transporters according to the related researches which discussed the equivalent mechanism of several transporters in material handling unit. Considering the batch size in the material handling unit of flexible flow shop is variable, these two divisional sub-problems have been described respectively by establishing their own open queueing network models with finite buffers.Secondly, as the open queueing network modeling with finite buffers has no certain product-form solution, state-space decomposition method has been used to get the approximate solution of the model. However, traditional state-space decomposition method can’t work out the solution to the open queueing network whose batch size in the material handling unit of flexible flow shop is variable. Based on the theory of discrete Markov process, along with the specific features of the open queueing network, the state-space decomposition method has been modified in this paper, to obtain the approximate solution of the model stated above. For the modified method, three kinds of nodes, the node of single arrival with bulk leaving (the buffer before the car), the node of bulk arrival with bulk leaving (the car) and the node of bulk arrival with single leaving (the buffer after the car), have been created and their state-space models with state balance equation functions have been built by redefinition of state variablebi.Thirdly, simulation models of open queueing network models for these two sub-problems have been built respectively to testify the effectiveness of the modified state-space decomposition method. And the conversion relationship of performance analysis before and after the equivalent mechanism of several transporters in material handling unit is validated through the simulation experiments by the simulation model above.Finally, the comparison experiments between simulation model and state-space decomposition method test and verify the availability of the method, and confirm the precision of the solution. Then the system performance of flexible flow shop with material handling unit has been analyzed based on the experiments stated above.In brief, using the theories of queueing network, queueing network models for the flexible flow shop with material handling unit have been built and approximate solutions for the model that have been created. Then an initial system performance analysis of the flexible flow shop with material handling unit has been made. The achievement and conclusion of this research can support the MTO managers on long-term development planning both in theory and decision.