供应商选择问题的不确定性模型及遗传算法(英文)

Application of Uncertain Optimal Model and Genetic Algorithms to Vendor Selection Problem

供应商选择问题是物流领域内的一个重要问题,其目标函数就是在包括送达时间、原料质量和服务水平等的约束下使总花费最小.传统的确定性模型取得了较好的效果,但是由于其约束条件的随机性和模糊性,需要应用不确定性模型来更准确地描述和给出最优解.论文应用了不确定性模型描述这类供应商选择问题,这个不确定性模型就是一类特殊的机会约束规划模型,此类机会约束可以转化为相应的等价类,这样不确定性模型就可以转化为确定性模型,然后设计了解决这类问题的遗传算法.通过实例计算表明,不确定性优化模型及遗传算是解决供应商选择等这类不确定性智能商业问题的有效办法,有广泛的应用前景.

The vendor selection problem is one of important problem in logistics, and it needs to minimize the aggregate price with constraints of the vendor such as price, delivery lead-time, material quality, service level, etc. The traditional certain models obtain some good results, but the vendor constraints are so stochastic and fuzzy that it needs uncertain models to describe and solve the problem. The paper applies uncertain model to describe the vendor selection problem, and the uncertain model is a class of special stochastic chance-constrained programming models. In our case the chance constraints can be converted to their respective deterministic equivalents, then the equivalent deterministic models could be obtained. Genetic algorithm has been designed for the problem. The experiments demonstrated the genetic algorithm and uncertain optimal models could be promising ways for uncertain intelligent business area as vender selection problem.