Metaheuristics for Inventory Control Policies with Random Deterioration Start, Preservation Investment and Pre-payment

In the real scenario, some items deteriorate over time but maintain their original quality for a while. However, it is not possible to estimate the deterioration start time (DST) for a specific item as many factors affect the degradation in the inventory system. To deal with uncertainty in the start of degradation, we investigate an inventory control policy for delayed degradable products, where the deterioration start time is a random variable and can be extended by making use of preservation investment. A price and stock level dependent demand is considered to formulate a lot-sizing policy. The supplier mandates that a fraction of the purchasing price is to be paid before the delivery by the retailer, and in exchange, the supplier offers a discount on the total price. To deal with a mixed-integer optimization problem associated with the inventory system, which is highly nonlinear, metaheuristic optimization approaches, viz., differential evolution (DE) and particle swarm optimization (PSO), have been employed. The best-suited metaheuristic algorithm is chosen by using ANOVA and convergence criteria. Numerical simulation and sensitivity analysis of relevant parameters, which have vital managerial applications, are carried out.