主从递阶多目标风险决策模型构建及算法研究  

Algorithm Study on Models of Multiple Objective Risk Decision under Principal and Subordinate Hierarch Decision-making

作　　者：王海政[1] 仝允桓[1] 

机构地区：[1]清华大学经济管理学院,北京100084

出　　处：《运筹与管理》2007年第1期1-8,共8页Operations Research and Management Science

基　　金：国家自然科学基金重点资助项目(70233001)

摘　　要：针对一类表现出主从递阶、多目标、风险决策等特点决策问题,文献回顾指出目前问题是缺乏一体化集成研究。为了研究该类问题,首先构建了主从递阶多目标风险决策模型。引入模糊随机模拟、多目标满意度计算、主从递阶博弈、基于粒子群算法的演化博弈求解技术,建立了具有逻辑关系的算法体系。案例研究表明算法是可行的,可为同类问题研究提供算法。None of the articles have ever addressed a decision-making problem with the charactensucs ot principal and subordinate hierarch decision, multi-objective decision, and risk decision to be simultaneously solved before. For the purpose of filling this vacancy, this paper constructs multiple objective risk decision models with principal and subordinate hierarch decision-making by fuzzy stochastic simulation technology, the multi-objective degree of satisfaction computation, principal and subordinate hierarch game decision method, and evolutionary game solution result is convincible after a case study. technology with particle swarm algorithm. The result is convincible after a case study.

关 键 词：主从递阶决策 多目标决策 风险决策 演化博弈 模糊随机模拟 粒子群算法 

分 类 号：C934[经济管理—管理学]