Closed-loop Supply Chain Network Design for E-Commerce Environment: an Integrated Framework Based on Mixed Integer Programming and Intelligent Optimization Algorithm

In the e-commerce environment, the design of closed-loop supply chain networks faces the challenges of spatio-temp oral heterogeneity of demand and complexity of return flow. Considering at this problem, this paper proposes an innovative framework integrating mixed integer programming and intelligent optimization algorithms. The framework accurately predicts dynamic requirements through the GCN-GRU model, combines distributed robust optimization to deal with uncertainties, and introduces a cooperative game mechanism to achieve fair cost sharing. Moreover, it adopts a two-stage adaptive optimization strategy. The initial network scheme is generated in the first stage, and the ALNS algorithm is dynamically adjusted in the second stage. The experimental results show that the average total cost of the IMIP-IOA model proposed in this paper is 680.4 thousand yuan under the scale of 200 nodes, which is 5-10% lower than that of the comparison algorithm. When the demand fluctuates 20%, the cost volatility is only 4.7%, and the solution success rate still maintains 95% in a 2000-node large-scale network, which significantly improves the solution quality, robustness, and scalability. This study provides an effective decision support tool for e-commerce closed-loop supply chain, and its architecture integrating prediction, optimization and game provides a new idea for the optimization of complex logistics system.