Multi Objective Quantum Evolutionary Algorithms for Global Supply Chain Optimization

Abstract

Global supply chain systems become complex with respect to globalization and the requirement of making fast decisions, sometimes with several conflicting objectives in terms of costs, time delivery, emissions, and resilience. Classical MOEA approaches, such as the genetic algorithm approach and the NSGA-II framework, have been used for such supply chain problems but are subject to premature convergence and poor scalability issues. The Quantum Evolutionary Algorithm (QEA) method is proposed for this study to combine the ideas of quantum computation with the concept of MOEA. The problem variables are represented in the form of qubits to perform parallel exploration of different solution candidates. QEA-based crossover and mutation operators can be used for the probabilistic variation of candidate solutions. The idea of Pareto front ranking helps preserve the optimal solutions. Benchmark datasets related to global supply chains were analyzed in this paper, and their performance was compared with classical MOEA algorithms. The QEA algorithm was found to perform better than its rivals in terms of hypervolume and convergence speed. Through ablation studies, the significant contribution of the quantum operator towards the enhancement of quality solutions and faster convergence rate has been proven. The study shows that QEA offers a realistic and efficient technique for dealing with the multi-objective problems associated with the complexities in global supply chains. For future research work, efforts will be made to scale-up the process using large quantum computers, apply the methodology to the dynamic nature of supply chains, and explore hybrid classical and quantum optimization techniques.