A multi-objective integrated optimisation model for facility location and order allocation problem in a two-level supply chain network

This study proposes a mixed-integer multi-objective integrated mathematical model solving facility location and order allocation optimisation problems simultaneously in a two-echelon supply chain network. The proposed problem is motivated by a factoryless concept and by providing a dynamic decision-making solution under a multi-period time horizon. Within the model, we also determine the optimal replenishment number of production facilities by the multi-objective functions. The multi-objective functions include minimisation of the total cost, rejected and late delivery units and, maximisation of the assessment score of the selected suppliers. The studied dynamic decision model is significant for the cost-efficient management of companies' supply chain networks. The mixed-integer mathematical model is developed by the LP-metric method and it is solved by the GAMS optimisation software. Due to the NP-hard structure of the problem, for large-scale instances, we utilise the Multi-Objective Particle Swarm Optimisation (MOPSO) and Multi-Objective Vibration Damping Optimisation (MOVDO) heuristic solution approaches. Numerical results show that, for large-scale problems, the MOPSO method performs better in Pareto solutions and decreases run times. However, the MOVDO method performs better regarding the Mean Ideal Distance and the Number of Solutions Cover surface criterion. The developed solution approach by this paper is a generic model which can be applied for any two-level network for simultaneous optimisation of supplier selection, location determination of facilities and their replenishment amounts.

Automated summary

This study proposes a mixed-integer multi-objective integrated mathematical model for solving facility location and order allocation optimization problems in a two-echelon supply chain network. The model considers multiple objective functions, including cost minimization, rejected and late delivery unit minimization, and maximization of the assessment score of selected suppliers. This study is significant for cost-efficient management of companies' supply chain networks.