近年来废弃电气电子设备的回收处理已成为全世界范围内亟待解决的环境问题。作为典型的逆向供应链环节，电子废弃物回收处理往往需要回收多种类的电子元器件，供需情况复杂多变，因而具备更高的环境不确定性与结构复杂性。另一方面，区别于其他可回收物，不同种类的电子废弃物需采用不同的处理工艺，产出对应的贵金属等资源。受回收品质量等因素影响其资源产出量也具有一定的不确定性。因此,在电子废弃物产生量和处理资源产出量不确定性的情况下，如何决策各处理中心的建设、产能与回收技术的投资，使得回收网络的处理能力和收益最大化，是电子废弃物逆向供应链管理中亟待解决的问题。 为探索在面对多重不确定性的情况下提升逆向供应链运营管理水平的方法，本文选取了电子废弃物的回收处理环节作为逆向供应链的代表，考虑了电子废弃物产生量和处理资源产出量的双重不确定性，采用分布式鲁棒优化的方法将电子废弃物回收处理网络的管理问题建模为包含投资阶段和运营阶段的两阶段随机规划模型。通过引入自适应性分布鲁棒重构，本文将原两阶段随机规划问题转换成可求解的二阶锥规划模型。为满足面向大规模回收处理网络规划问题的求解需要，本文提出了外逼近求解算法，显著提高了模型的求解效率。 在不确定性应对方面，冗余策略和流程柔性策略作为正向供应链中应对不确定性的典型策略得到了广泛研究，但面向逆向供应链管理的不确定性应对策略研究仍不够充分，特别是对于混合不确定性应对策略的研究较为匮乏。因此，本文针对产能冗余和流程柔性设计了两类混合策略，探索了产能冗余和流程柔性策略的交互影响。通过开展基于国内电子废弃物回收数据的数值实验，本文发现了同时使用产能冗余和流程柔性策略产生冲突的情况，有助于提高对冗余策略、流程柔性策略及混合不确定性应对策略的理解，为考虑多重不确定性的电子废弃物回收处理网络的规划运营提供了指导建议。

Recently Waste Electrical and Electronic Equipment has become urgent environmental issues around the world. As a typical type of reverse supply chain, in e-waste recycling generally various types of electronic components are recycled with complex and changeable relations between supply and demand, thus resulting higher environment uncertainty and structural complexity. On the one hand, e-waste differs from other recyclables in that different types of e-waste require corresponding processing technology to obtain resource yields like precious metals. The resource yields are also uncertain affected by factors such as the quality of e-waste. Therefore, considering the uncertainty of e-waste generation and resource yields, the optimization for investment of recycling center, capacity and the choice of recyling technologies is addressed urgently in the e-waste reverse supply chain management to maximize the service level and recycling yields. To improve the operation of reverse supply chain confronted with multi uncertainty, this research chooses the recycling link of e-waste reverse supply chain with uncertainties of both e-waste generation and resource yields as study object. The e-waste recycling process is formulated as a two-stage stochastic programming model including the investment and operation stage. Uncertainties of both wastes returns and resource yields are considered and handled by a distributionally robust optimization method, then a computational tractable second-order cone programming model is obtained using an adaptive robust reformulation. Furthermore, an outer approximation algorithm is proposed to meet the needs of solving large-scale network planning problems and shows superior performance in computation efficiency. To meet the challenge of supply chain uncertainty, redundancy and process flexibility strategies have been widely studied as typical risk mitigation stratigies in forward supply chain, while the research on strategies in reverse supply chain management is still insufficient. Specially, the research on hybrid uncertainty mitigation strategies is rare. Therefore, this study designs two types of hybrid strategies for capacity redundancy and process flexibility, and explores their coupling effects. By implementing numerical experiments based on real data, this research finds conflict situations of using capacity redundancy and process flexibility strategies together, contributes to improve the understanding of redundancy and process flexibility strategies as well as their coupling effect, and gives guidelines for the use of hybrid strategies in e-waste recycling reverse supply chain management.