Aimed at some problems that the resource allocation and task with high priority are distributed with assets first, and task with low priority fails to get enough resource to accomplish a task, a matching strategy is proposed. This new method has two parts. One is a Nbest algorithm based on decision space partitioning theory, and this can provide decision makers with several alternative plans to overcome the limit of a single plan. The other is a tradeoff strategy combined with N best plans to get the best solution, thus realizing the whole situation's optimization. The experiment results show that the Nbest based strategy can improve the whole situation's task accuracy.