公共部門多目標投資計畫選擇之研究

Abstract

Numerous quantitative models for project selection have been developed in recent twenty year. Of these models, the involement of multiple objectives to select projects is a common trend. The problem of selecting investment projects can generally be decomposed into two stages: first to select a set of projects which can achieve a given multiple objectives from all feasible candidate projects, and secondly to schedule those chosen projects. In this paper, a max-min multiobjective integer programming model is discussed and a numerical example of goods distribution center is illustrated for this model. The optimal set of projects achieving the given objectives over the planning period will be selected from the model. Two spatial methods and their solution algorithms, primal spatial method and dual spatial method for solving multi-objective integer programming. In the max-min integer pro?gramming, a benefit-cost ratio is used as an important index to determine the efficiency of project selection, subject to some resource constraints. The numerical example of goods distribution center suggests that the same four projects are selected from eight investment projects by two spatial methods.