基于共谋合作机制的补偿滑轮组变幅机构多目标博弈设计  

Multi-objective Optimization Design Method Based on Coalition Cooperative Game Theory and Its Application in Structural Design

作　　者：李碧研[1] 谢能刚[1] 孟瑞[1] 

机构地区：[1]安徽工业大学机械工程学院,马鞍山243002

出　　处：《机械科学与技术》2011年第10期1669-1672,共4页Mechanical Science and Technology for Aerospace Engineering

基　　金：教育部新世纪优秀人才计划(NCET-07-0003)资助

摘　　要：综合考虑稳定性和经济性等多目标设计要求,建立补偿滑轮组变幅机构的多目标优化设计模型。提出多目标共谋合作博弈设计方法,将3个设计目标视为3个博弈方,将设计变量集合分解为各博弈方拥有的策略空间。根据共谋机制,建立博弈得益函数,各博弈方分别以自身博弈得益函数为目标,在各自的策略空间中进行单目标优化,获得本博弈方对其余博弈方的最佳对策,所有博弈方的最佳对策形成一轮博弈的策略组合,并根据收敛判别,通过多轮博弈,获得最终的博弈解。对实际算例进行了三目标博弈设计,计算结果显示了共谋合作博弈方法的有效性。Considering the requirement of multi-objective design comprehensively, such as the stability and econo- my, the multi-objective optimization model of luff mechanism of compensative sheave block was built and a multiobjective optimization design method was proposed based on coalition cooperative game theory where the three design goals were treated as three game players. The design variables were divided into strategy space of game players and revenue function was constructed based on conspiracy mechanism. Each game player took its own revenue function as target with single objective optimization in their own strategy space in order to get its own best strategy. The best strategy of all players form a combination of game strategy and the optimal solution can be obtained through several game rounds. Example results show the effectiveness of coalition cooperative game method .

关 键 词：共谋合作博弈 补偿滑轮组变幅机构 多目标优化 模糊聚类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] O221.6[自动化与计算机技术—计算机科学与技术]