Abstract: By characterizing the return rates as fuzzy variables，this paper proposes the chance criterion models for optimization decision-making of loan portfolio of commercial banks，i.e.，possibility criterion model，necessity criterion model and credibility criterion model.For the case where the loan return rates are special triangular fuzzy variables，the crisp equivalents of the models are given and can be solved with the traditional methods.The hybrid optimization algorithm is employed to solve the models where the membership functions of loan return rates are complex.The algorithm integrates fuzzy simulation，neural network，genetic algorithm and simultaneous perturbation stochastic approximation algorithm.Numerical examples illustrate the effectiveness of the models and the algorithm.

Key words: loan portfolio, chance criterion, fuzzy variable, fuzzy simulation, genetic algorithm（GA）

摘要： 通过把贷款的收益率刻画为模糊变量，提出了商业银行贷款组合优化决策的机会准则模型，即可能性准则模型、必要性准则模型和可信性准则模型。对于贷款收益率是特殊的三角模糊变量的情况，给出了模型的清晰等价类，这些等价类可以用传统的方法进行求解。对于贷款收益率的隶属函数比较复杂的情况，应用集成模糊模拟、神经网络、遗传算法和同步扰动随机逼近算法的混合优化算法求解模型。数值算例验证了模型和算法的有效性。

关键词: 贷款组合, 机会准则, 模糊变量, 模糊模拟, 遗传算法