Optimizing the inventory control decisions under multiple constraints for deteriorating products: An application of meta-heuristic algorithms

This study aims to investigate the influence of known price increases on the inventory model regarding both uniform and an exponential distribution of replenishment intervals with the partial backorder. It examines the optimization of inventory control decisions for deteriorating products considering a known price increase, probabilistic replenishment interval, warehouse capacity constraint, and partial back-ordering.

Design/methodology/approach:

 To obtain the specific inventory order quantity, the problem has been modeled in such a way that the total cost savings function is obtained from the differences in the optimal order policy for both special and regular orders. The two situations discussed in this study are: i) unconstrained problem modeling, and ii) constrained problem. Some computational experiments have been performed to examine the effects of various parameters on cost savings performance. For the constrained problem, Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) have been used and their results have been compared in terms of the cost savings values and computation time.

Findings indicated that for the constrained problem, GA has a better performance than PSO. Accordingly, for an unconstrained problem, by using the derivative of the profit function and performing sensitivity analysis, the influence of parameters such as demand, price, holding cost after the price increase, λ in exponential distribution, length of periods in uniform distribution, and deterioration rate on the decision variables including order quantity and the profit were.

 obtained,Practical implications:

 The model’s generated policy is more effective and profitable for retailers when demand and deterioration rate are higher and replenishment periods are decreased.

Originality/value: 

This study completes the previous inventory control models that were under the policy of known price increase and is closer to the real environment by utilizing deteriorating items, capacity constraints, and meta-heuristic approaches.