Optimizing Inventory Level in Multi-Period and Classification Items Problems

Abstract

This article examines how simulation can manage demand distribution for different products to make an expected profit. When demand is unpredictable and variable due to seasonality within the week, and significant uncertainty, ordering perishable items has become challenging. Ordering fresh products with a long shelf life is one approach to prevent wastage and provide a high level of service, but the quality suffers as a result. Based on the assumption that order size will follow a log-normal distribution with unknown parameters, we assumed the distribution would follow a log-normal distribution. Retail stores should implement policies for continuous review of inventory levels every day to prevent shortages. These policies can be easily implemented. For a number of inventory products, the metaheuristics methods are also used to determine the ideal order amount and reorder point. To maximize long-term average profit in our situation, several control criteria must be established to account for predictable variability. Relevant data are the profit margin and the loss due to scrap items. This research handles stochastic demand for different items and a continuous-review period inventory management model to maximize its expected profit. The framework of an integer nonlinear programming model is presented based on random demands. In addition, grey wolf optimization (GWO) and particle swarm optimization (PSO) approaches are employed to realize the optimal quantity order and reorder point. Optimal solutions from GWO and PSO are also verified through numerical results. Besides, ABC analysis, enchanted by machine learning, is applied to classify items based on their importance to get higher revenue. The numerical analysis and results are demonstrated on multi-objective inventory optimization in each of the four scenarios. With a decision-making scheme powered by optimization algorithms, novel inventory management software can learn an ever-fluctuating production flow and anticipate the need for changes in the real market.