Multiattribute Project Portfolio Selection Problems Based On SMAA

Nowadays, with the rapid development of economy and society, the decision environment change fast. How to choose the most effective project portfolio from many seemingly attractive projects is an important task in front of managers in such fast changing decision environment. If the managers choose the right project portfolio, it can bring sustained, healthy and stable development for the organization. If you select the wrong project portfolio, it will waste valuable resources of the organization, and hinder the progress of the organization, even endanger the future of the organization. In this paper, we focus on the problem of multi-attribute project portfolio selection problem. In a typical multi-attribute project portfolio selection problem, the decision-maker should select select the optimal portfolio from large number of candidate project according to some attributes, under the resource constrains. However, this is not a simple task since the following reasons.1, the decision makers should assess the amount of organizational resource comprehensively. The amount of organizational resources not only decided which projects can be under construction, but also decided how many projects can be done simultaneously. However, as the resource types are multiple, the existing forms of resource are various and the resource levels of organization are changing fast, it is difficult for the decision makers to assess the accurate amount of resource owned by organization.2. In a rapidly changing decision-making environment, it is difficult to know the precise performance of multiple attributes.3. The preference information is hard to know. Different decision makers may have different preference information.4. In the multi-period project portfolio selection problem, the decision makers not only consider which project portfolio is optimal, but also consider the best start time for each project.The multiple attribute project portfolio selection problems, in view of its difficulty and important meaning, have becoming one of the hot problems in the operation field. However, the existing research methods cannot solve the above problems perfectly. In this paper, we propose a series of models based on stochastic multi-attribute accessibility analysis, which can solve the above problem successfully.This paper structured as follows. The first chapter is the introduction. We mainly introduce research background, the existing research literature, research objectives, research methods and the structure of the thesis. In chapter2, we introduce the basic theory tools of the paper. The main research methods used in this thesis are the SMAA methods. In chapter3, we focus on the study of single period project portfolio selection problem with uncertain decision environment. In this chapter, we propose two models which are SMAA-PO and SMAA-P02. The SMAA-PO model can be used to solve the single-period project portfolio problem with independent project. While the SMAA-P02model can be used to solve the single-period project portfolio problem with attribute interdependence. In chapter4, we focus on the study of multi-period project portfolio selection and scheduling problem with uncertain decision environment. In this chapter, we also present two models. The first model is used to solve the multi-period project portfolio selection problem and the second model is aimed to solve the project portfolio selction and scheduling problem. In chapter5, we focus on the multi-attribute project portfolio selection and adjustment problem. In this section, we consider the adjustment strategies of the existing projects and the selection of candidate projects simultaneously. The method proposed in this section can help the decision makers to select the best resource allocating strategies and gain the maximization of comprehensive benefit of organization. The last chapter is the conclusion. In this chapter, we give the main conclusions of this paper and the possible research directions in the future.The main contributions of this thesis are:(1). we employ random values and rank levels information to capture the uncertain attribute values. We use weight distribution to deal with completely unknown or partly unknown weight information.(2) We propose SMAA-based model for the single-period multi-attribute project portfolio selection problem with uncertain attribute values, unknown weight information and project interdependence.(3) We present useful methods for the multi-period project portfolio selection and scheduling problems. Compared with the existing related researches, the presented methods fully considers the uncertain project duration, changing resource levels, multiple types of resource, unknown weight information and uncertain attribute values.(4) This thesis study the multi-attribute project portfolio adjusting and selection problems and propose solution methods based on acceptability analysis. The presented methods consider the adjustment strategies of the existing projects and the selection of candidate projects simultaneously.(5) We have extended the SMAA methods into the field of project portfolio selection. The existing SMAA methods have been improved in order to capture the actual needs of the project portfolio selection problems.