基于情景生成的银行卡网络资金配置鲁棒优化模型

A Robust Optimization Model for Fund Allocation of Bankcard Network Based on Scenario Generation

运用鲁棒优化理论和情景生成方法,研究了银行卡网络资金配置问题.在对银行卡网络资金需求分析的基础上,运用鲁棒优化理论建立了银行卡网络资金配置鲁棒优化模型.该模型为一个多目标规划问题,满足诸如尽可能全天候满足持卡人需求、系统的总收益最大、网络的鲁棒性等多个相互冲突目标.以我国金融市场为背景,通过仿真-优化-聚类情景生成方法描述银行卡网络资金的不确定需求,并采用遗传算法对模型求解.研究结果表明,银行卡网络资金配置鲁棒优化模型给出的解,能实现银行卡网络鲁棒性.

This paper studies the fund allocation problem in the environment of China UnionPay Network.Based on analyzing the fund demands of bankcard in China UnionPay Network,we establish a robust optimization model for fund allocation of bankcard network by using robust optimization.In this model,we describe the uncertain fund demands of bankcards by a hybrid scenario generation based on simulation and optimization.Moreover,we simulate the model by genetic algorithm optimization toolbox in the Matlab and obtain the cash and account funds which banks in the network should prepare for satisfying the uncertain fund demands.The result shows that the cash and account funds obtained from the model can satisfy the uncertain fund demands of bankcards.It ensures both the stability and the maximization of the total profit of the network.