A New Inventory Control Approach For Considering Customer Classes In An Integrated Supply Chain Management

TL;DR

This paper introduces a novel integrated supply chain model that accounts for customer differentiation, stochastic demands, and multi-level capacity constraints, employing exact and heuristic methods for optimization.

Contribution

It develops a new multi-product, single-period supply chain model considering customer classes, stochastic demand, and transportation variability, with solution approaches for small and large instances.

Findings

Hybrid meta-heuristics improve solution quality for large instances.

ANOVA confirms the effectiveness of the proposed algorithms.

Sensitivity analysis reveals key factors affecting computational time and costs.

Abstract

Supply chain management is an integrated approach for planning and controlling materials, information, and finances as they move in a process which begins from suppliers and ends with customers in forward approach. As distribution network planning is strategically done, the related decisions should be optimized. This supply chain planning involves transportation, the location of facilities, and inventory control decisions. This study is a new approach for considering customers' differentiation in an integrated model to location-allocation and inventory control supply chain. The proposed model is multi-product, single-period and with stochastic demands. Additionally, warehouses have multilevel capacity limitation. For more reality, the probability of transportation through different vehicles, different transportation capacity, and transportation costs are also taken into the consideration. The customers are divided into two strategic and nonstrategic groups by adopting the critical level policy. The exact calculation method is employed for small scale instances while hybrid meta-heuristic algorithms (Genetic and Simulated Annealing) developed for real samples. Efficiency and quality of solutions are examined via the ANOVA to report proper and near optimum solution. Finally, sensitivity analysis is carried out for different instances to evaluate the effect of different indexes on the duration of CPU time and values of the objective function.