APPROXIMATION APPROACH FOR THE PERFORMANCE ANALYSIS OF PRODUCTION LINES UNDER A KANBAN DISCIPLINE

In a JIT production environ:ment, pull-type control systems are usuall y implemented to reduce the lead time and to synchronize the productio n rate with the demand rate. The design of an effective production con trol system requires easy and reliable approaches for evaluating the e ffect of the major system parameters on its performance. In this paper we model the kanban control system of a manufacturing line as a Marko v process, In addition, we develop an approximation approach to the mo del that permits reliable evaluation of manufacturing system performan ce in terms of throughput time and work-in-process. The validation of the approach is implemented by comparing the exact results (based on d iscrete event simulation) and the approximated results. A good approxi mation is observed for a large range of conditions.