Robust optimization of multi-objective multi-cycle remanufacturing supply chain network considering incentive compatibility theory under improved algorithm

Abstract

In recent years, industrial production and transportation have generated a significant amount of carbon emissions, leading to a heightened focus on environmental resource conservation. Concurrently, with the development of internet technologies, an increasing number of enterprises are leveraging big data and artificial intelligence for production and logistics. Within this context, the challenge of how enterprises can utilize big data services to reduce costs while minimizing carbon emissions has become a pressing concern. Meanwhile, remanufacturing, as a sustainable solution, has garnered substantial attention. However, remanufacturing supply chain networks become exceptionally complex and challenging when faced with uncertainties in demand and recovery quantities, necessitating government intervention. Thus, to examine the impact of government intervention on the optimization of green remanufacturing supply chain networks, this paper, based on incentive compatibility theory, establishes an optimization function with objectives of minimizing total cost, reducing carbon emissions, and maximizing return on data investment in scenarios marked by uncertain market demand and recovery quantities. It constructs multi-objective fuzzy optimization models for various scenarios. The paper employs improved robust optimization to address uncertainties related to demand and recovery quantities. It also utilizes the Improved Multi-Objective Genetic Algorithm (IMOGA) to address multi-objective optimization, addressing the limitations of Genetic Algorithms (GA) and Multi-Community particle swarm optimization (MPSO). Case study results indicate that robust optimization effectively handles uncertainties in the model, aligning better with real-world conditions. High-tech enterprises often prioritize economic objectives, and government intervention measures can compensate for their limited carbon reduction capabilities. In most cases, the improved multi-objective genetic algorithm (IMOGA) outperforms MPSO and GA in terms of performance. In the context of the evolving internet landscape, high-tech enterprises should harness big data and artificial intelligence to establish green remanufacturing supply chain networks. Simultaneously, government guidance to reduce carbon emissions is instrumental in enhancing corporate competitiveness and industrial upgrading.