LINGUISTIC FUZZY VARIABLES AS ANALYSIS TOOL IN INVENTORY MANAGEMENT

Abstract

Decisions associated with supply and inventory control in industrial production planning process is frequently made on the basis of vague and imprecise information about factors that affect inventory costs. In order to minimize the total expected inventory cost and timely satisfy market demand in conditions of uncertainty and insecurity, the classic inventory control methodologies can be upgraded with the control systems based on fuzzy logic. In this paper, as convenient framework for dealing with vagueness in inventory decision making, we introduced the use of linguistic fuzzy variables processed by fuzzy if-then rules in the fuzzification and defuzzification model. In proposed fuzzy inventory model, there are two input variables: “demand value for a product“ and “quantity on hand parts“ needed to build a product. These input variables are presented by fuzzy sets containing five terms. In addition to the input variables, model includes the if-then rules with one output variable – the „inventory action“ which suggests reordering of parts, reducing the number of the already existing parts or no action at that time. The output variable is a fuzzy set that contains seven terms. All of these linguistic fuzzy variables are presented by fuzzy triangular numbers and parts of fuzzy trapezoidal numbers. For defuzzification (in order to obtain a crisp number) we used the mean of maximum method (MMM). Finally, the resulting value (in percentage) is translated into a corresponding inventory action. Generally, this paper emphasizes the usability of fuzzy inventory models in dealing with the uncertainty of demand and overall assessment of inventory costs based on the experience and subjective evaluation of managers.