Optimizing lot sizing model for perishable bread products using genetic algorithm

This research addresses order planning challenges related to perishable products, using bread products as a case study. The problem is how to effi­ci­ently manage the various bread products ordered by diverse customers, which requires distributors to determine the optimal number of products to order from suppliers. This study aims to formulate the problem as a lot-sizing model, considering various factors, including customer demand, in­ven­tory constraints, ordering capacity, return rate, and defect rate, to achieve a near or optimal solution, Therefore determining the optimal order quantity to reduce the total ordering cost becomes a challenge in this study. However, most lot sizing problems are combinatorial and difficult to solve. Thus, this study uses the Genetic Algorithm (GA) as the main method to solve the lot sizing model and determine the optimal number of bread products to order. With GA, experiments have been conducted by combining the values of population, crossover, mutation, and generation parameters to maximize the feasibility value that represents the minimal total cost. The results obtained from the application of GA demonstrate its effectiveness in generating near or optimal solutions while also showing fast computational performance. By utilizing GA, distributors can effectively minimize wastage arising from expired or perishable products while simultaneously meeting customer demand more efficiently. As such, this research makes a significant contri­bution to the development of more effective and intelligent decision-making strategies in the domain of perishable products in bread distribution.