一种求解多目标整数规划的新方法

A New Method of Solving Multi-objective Integer Programming

在求解多目标整数规划的加权方法中,交互式方法虽然考虑了决策者的偏好,但计算量较大;非交互式方法计算量虽小,但没有考虑决策者的偏好。因此,提出了基于层次分析法的权系数确定方法,并构造了相应的算法,该算法不仅计算量小,而且考虑了决策者的偏好,并针对数据不平衡问题,对算法进行了改进,给出的算例表明算法是有效的。

In the weighting method of solving multi-objective integer programming, interactive method takes the preference of decision-maker into consideration, but needs large complexity. Noninteractive method takes the reverse approach. This paper presents a new weighting method based on analytic hierarchy process （AHP）. The corresponding algorithm gets a balance between the preference of decision-maker and computational complexity. The algorithm is improved to deal with data nonbalance. At last, an example shows that the algorithm is efficient .