摘要
A decision support model with stochastic multiobjective functions and constraints and its solution procedure are presented for selecting R&D projects, in which the uncertainty of project evaluation and selection, the interactions of technique, resource and benefit among projects, and the experience, knowledge and preference of R&D managers are considered. The statistical results of stochastic factors representing the benefit contributions and the three kinds of interactions corresponding to each objective among projects are obtained by the Delphi method and QS/NI decision process for developing the model. In the decision analysis solution procedure, experiences and preferences of R&D managers to make decisions are considered by means of aspiration goal levels(AGL), probabilities to reach AGL called satisfied degrees, and probabilities to satisfy the corresponding constraints called feasible degrees, under which a project selection plan can be proposed by the decision analyst or computer. The utility of this model and solution procedure are demonstrated by a real world case example in R&D program of strength and vibration of aircraft turbine engines. which involves 23 project candidates. The evalutions of benefit contributions and interactions of projects are made by 41 experts in the field.
A decision support model with stochastic multiobjective functions and constraints and its solution procedure are presented for selecting R&D projects, in which the uncertainty of project evaluation and selection, the interactions of technique, resource and benefit among projects, and the experience, knowledge and preference of R&D managers are considered. The statistical results of stochastic factors representing the benefit contributions and the three kinds of interactions corresponding to each objective among projects are obtained by the Delphi method and QS/NI decision process for developing the model. In the decision analysis solution procedure, experiences and preferences of R&D managers to make decisions are considered by means of aspiration goal levels(AGL), probabilities to reach AGL called satisfied degrees, and probabilities to satisfy the corresponding constraints called feasible degrees, under which a project selection plan can be proposed by the decision analyst or computer. The utility of this model and solution procedure are demonstrated by a real world case example in R&D program of strength and vibration of aircraft turbine engines. which involves 23 project candidates. The evalutions of benefit contributions and interactions of projects are made by 41 experts in the field.
