中文摘要：

针对现有研究大都仅对单个目标进行优化的现状，提出一种优化决策模型，即将企业信用风险迁移引入到贷款收益率的计算中，以限制资产数目、决策变量上下界等为约束条件，建立组合投资的收益最大、同时方差风险最小的优化决策模型。该模型是一个混合0—1变量的多目标规划问题，提出求解该模型的一种自适应改变惯性权重的离散粒子群算法，数值结果表明该算法是有效的，模型是合理的。

英文摘要：

This paper transfers the corporate credit risk into the calculation of loan return. We set up a loan＇s portfolio hybrid O--1 multi--objective optimization model, in which the maximum portfolio profit and the minimum of risk of banks are taken as objective functions, meanwhile, the number of asset and the upper and lower limits are taken as constrains. A new discrete Particle Swarm algorithm with self adaptable changing inertia weight is presented to solve the problem. It is shown by the numerical result that the algorithm is effective and the given model is reasonable.