A Realistic Approach for Modeling Stochastic Lead Time Distributions

Abstract This paper presents some techniques for the realistic treatment of stochastic lead-time-demands in inventory models. First, we suggest the use of the first four moments to describe the diversified distribution forms of the lead times and the daily demands; and formulas for deriving a lead-time-demand's first four moments are presented. Next, we demonstrate the use of a lead-time-demand's first four moments in conjunction with the Johnson et al.'s tables to obtain various probability estimates. Thirdly, we discuss the use of the versatile four-parameter Schmeiser-Deutsch curves to fit a lead-time-demand distribution. The computational advantages of using the fitted Schmeiser-Deutsch curves in solving inventory models are then illustrated.