基于愿景满意度函数的混合型多属性决策方法研究

The complex multiple attribute decision making problem always faced with heterogeneous Information. In this kind of decision problem, considering aspiration can make the decision making more scientific and flexible, the recent research of this field based on prospect theory which focus on only one kind of the decision making information and less consider the influence of the risk preference to the aspiration satisfaction. So, it is necessary to study how to make the decision with aspiration satisfaction of different risk preference in the case of not unified the heterogeneous information. This project studies on hybrid multiple attribute decision making based on satisfaction function of attribute aspiration. Firstly, proposed a method of hybrid multiple attribute decision making considering attribute aspiration; Secondly, studies on hybrid multiple attribute decision making considering both the individual and collaborative attribute aspiration; finally, it is applied to solve the interactive hybrid multi-attribute decision making problem considering the rationality of the aspiration. There are two innovation and characteristics of this work: one is making the satisfaction function of attribute aspiration by combination attribute score, utility function and risk preference, and this function can be flexibly combined according to the different attributes, aspiration, risk preferences, and better reflect the effects of risk preference on aspiration satisfaction; the other is design the interactive algorithm by considering the reasonable interval of the individual aspiration and the acceptable deviation of the multi person aspiration. The research will complement and perfect the hybrid multi-attribute decision making theory and method.

复杂决策问题，常常是属性值包含多种数据形式的混合型多属性决策问题。在这一类决策问题中，考虑愿景可以提高决策的科学性和柔性，该领域研究主要基于前景价值函数，但是该函数不能很好的体现风险偏好、效用曲线随属性和愿景的变化而改变，也不适用于混合信息决策问题。所以，如何不统一信息模式，建立愿景满意度函数来体现风险偏好、效用曲线和愿景的相互影响有待解决。本项目旨在通过建立愿景满意度函数，解决考虑愿景的混合型多属性决策问题；扩展到考虑个体与协同愿景的情形；并考虑愿景合理性提出交互混合型多属性决策方法。主要特色在于：一是将愿景、属性得分、效用函数和风险偏好结合以构建愿景满意度函数，该函数可以根据属性、愿景、风险偏好的不同进行灵活组合，较好反映风险偏好对愿景满意度的影响；二是同时考虑单人愿景的合理区间和多人愿景的可接受偏差来设计调整愿景的交互算法。本研究将补充和完善考虑愿景的混合型多属性决策理论和方法。

本项目研究基于愿景满意度函数的混合多属性决策方法，具有重要的理论意义和实际意义。通过三年的研究工作，取得了较为丰硕的研究成果，其主要体现在以下三个方面：进行了考虑属性愿景混合多属性决策方法的基础性问题研究，提出考虑愿景的混合型多属性决策方法；提出了考虑个体与协同愿景的混合型多属性决策方法；提出了考虑愿景合理性的交互混合型多属性决策方法，并把所提出的方法应用于解决航空项目管理、医疗管理、风险投资和酒店管理等实际的决策问题。通过本项目的研究，较好地完成了预期制定的研究计划，达到了预期的研究目标。围绕本项目研究方向发表论文17篇，在电子工业出版社出版专著1部。在《Computers & Industrial Engineering》、《International Transaction in Operational Research》、《Applied Soft Computing》、《Soft Computing》、《International Journal of Fuzzy Systems》、《Journal of Intelligent & Fuzzy Systems》等国际重要期刊发表论文10篇，在《中国管理科学》、《系统工程理论与实践》、《管理评论》等国内重要期刊中发表论文7篇。本项目研究工作的完成，有益于对进一步发展和完善考虑愿景的混合多属性决策理论和方法体系。