资源分配问题两层优化分析的元模型方法

资源分配问题有着广泛应用,它的经典数学模型是线性规划问题。实际应用中资源的拥有量允许在一定范围内变化,资源分配优化所追求的应当是单位资源成本上的利润最大化。针对这一问题,提出了两层优化分析模型,并利用元模型给出了求解的方法。用二阶多项式元模型代替低层的较复杂的优化分析模型,将它嵌入到高层模型,可以实现快速可行的决策支持。数值实验结果表明了该方法的有效性。

Metamodel Method for Two-level Optimality Analysis of Resource Allocating Problem

The resource allocating problem is classically modeled as a linear programming problem and has wide applications in many fields.In practice,the amount of resource should be able to be adjusted inside certain range,and hence the manager would expect to acquire the maximal profit per unit of resources by varying the resources in an admissible set.A two-level optimality analysis model was proposed and metamodel methods were applied to solve this problem.The second-order polynomial regression metamodels were used as surrogate models for complicated optimality analysis models in the lower level and were embed to the higher level model.A rapid and feasible decision support tool for the manager was provided.Numerical experiments show the efficiency of this method.