Genetic Algorithm to solve Inventory problems in SMEs

Abstract

The objective of this study is to minimize the costs of the sales of the “Candies” store due to the accumulation of products in stock. This problem arises from the realization of constant fortnightly orders for fixed quantities, without taking into account the variability of demand. The alternative is then proposed to implement a Genetic Algorithm (GA) in Microsoft Excel using Visual Basic for Applications and Macros to carry out a projection corresponding to the inventory of the year 2020, in which the number of units for which each replenishment must be made is determined. . For the application of the algorithm, the historical data of the demand of the previous three years are taken: 2017, 2018 and 2019; It is worth mentioning that Monte Carlo simulation is applied to generate the data for the second semester of 2019. As a result, a reduction of approximately 100 products and more than $ 70 dollars was obtained with respect to previous years, which means an increase of 18% in the annual benefits of the products. In conclusion, the projection generated with the genetic algorithm provides a significant reduction in sales costs, since it reduces the amount of products that are usually wasted given their expiration before being dispatched. Automating AG deployment using Macros and Visual Basic reduces execution times and computational effort.