A hybrid machine learning solution for redesigning sustainable circular energy supply chains

Sustainability development goals require decision-makers to incorporate social and environmental indicators in their economic models using innovative solutions, such as a sustainable circular economy. This paper presents an innovative integrated production and logistic model for a circular economy using multi-objective optimization. Empirical data includes a renewable energy supply chain. We assess the sustainability performance of the proposed decision-making model by simultaneously considering production and logistics costs, carbon emissions, and the number of jobs created. The case study is optimized with an exact method, and a hybrid machine-learning algorithm solves large-scale numerical examples. The paper's main contributions include movable manufacturers, uncertain parameters, a hybrid machine learning algorithm, and empirical data in the proposed decision-making model. The findings show that a moveable facility can substantially decrease total cost and carbon emissions. Sensitivity analysis shows that changes in moveable capacity and percent yield considerably impact the objectives. Findings show that decision-makers can achieve cost parity with fossil-based sources when employing circular supply chain management.