Multi-commodity supply chain network design problem considering production cost and inventory holding cost

Abstract

This study works on an extension of mathematical optimization model by adding more cost components which are production cost at plant and holding cost at supplier and distribution center for the purpose of evaluating the final result and the impact of said costs. To achieve this, one potential existing model of cost minimization is chosen alongside metaheuristics algorithm named Particle Swarm Optimization with multiple social learning terms (GLNPSO) which are then applied to solve the problem. The experiments are conducted based on the data set and test problems from the benchmarks. Then the comparison of optimization models, before and after inclusion of two cost elements, is determined. This comparison stage attempts to illustrate the increasing amount of total cost/solution quality before and after cost inclusion, additionally to observe the coefficient of variation (CV) and average computational time from five replications of each test case problem. Following the benchmarks, the second comparison stage is essentially implemented by re-comparing the two variants of allocation scheme sharing the same proposed Multi-Commodity Supply Chain Network Design (MCSCND) mathematical model. The re-comparison would show whether decision making result remains the same or different from benchmark. Finally, the impact of two cost components is identified and significant to include in mathematical formulation model of location-allocation problem of MCSCND.