基于影响图对策的仿真元模型构造及使用方法
|
摘要
针对目前仿真元模型的不足,引入动态影响图对策到作为仿真元模型,提出了动态影响图对策构造步骤及使用方法。仿真状态、决策方案和决策目标分别用机会结点、决策结点和效用结点表示,对手的行为采用决策者相似的决策结点进行建模。构造基于影响图对策元模型的步骤包括选择变量和确定仿真数据、确定影响图对策的拓扑结构和概率表以及检验拓扑结构和概率表。通过空战仿真实例,表明该元模型在仿真的时间演变、假设分析和最优决策三方面的应用效果较好,克服了当前元模型的不足。
To overcome the shortcomings of the meta-model, the dynamic influence diagram game is used as simulation meta-model. A construction process and the utilization for structured dynamic influence diagram game(DIDG) are presented. The necessary simulation data are gathered for the construction of a DIDG that presents the changes in simulation state with chance nodes. The decision alternatives and objectives of the decision problem are included in the DIDG as decision and utility nodes. The adversary's action is represented by an explicit decision node in which case an influence diagram game results. The constructed DIDG is solved to find the optimal both sides' decision alternatives that maximize the expected value of the utility function. The DIDG is also used for studying the evolution of the simulation, conducting what-if analyses, and solving optimal strategies of the game as well as studying the strategies' effects on the evolution of simulation. The simulation results show the construction of DIDG from simulation data and their utilization in the analysis of decision problems are feasible and effective.
To overcome the shortcomings of the meta-model, the dynamic influence diagram game is used as simulation meta-model. A construction process and the utilization for structured dynamic influence diagram game(DIDG) are presented. The necessary simulation data are gathered for the construction of a DIDG that presents the changes in simulation state with chance nodes. The decision alternatives and objectives of the decision problem are included in the DIDG as decision and utility nodes. The adversary's action is represented by an explicit decision node in which case an influence diagram game results. The constructed DIDG is solved to find the optimal both sides' decision alternatives that maximize the expected value of the utility function. The DIDG is also used for studying the evolution of the simulation, conducting what-if analyses, and solving optimal strategies of the game as well as studying the strategies' effects on the evolution of simulation. The simulation results show the construction of DIDG from simulation data and their utilization in the analysis of decision problems are feasible and effective.
引文
[1]J P C Kleijnen.Regression and Kriging metamodels with their experimental designs in simulation[J].European Journal of Operational Research(S0377-2217),2016,248(2):93-96.
[2]王广彦,贾希胜,李经中,等.遗传算法的经验公式仿真元模型的构建方法[J].火力与指挥控制,2011,36(10):119-123.WANG Guangyan,JIA Xisheng,LI Jingzhong,et al.The Study of Experiential Formula Simulation Metamodel Based on Genetic Algorithm[J].Fire Control&Command Control,2011,36(10):119-123.
[3]J P C Kleijnen.Kriging metamodeling in simulation:A review[J].European Journal of Operational Research(S0377-2217),2013,230(3):707-716.
[4]C S Longstreet.A meta-model for developing simulation games in higher education and professional development training[C].International Conference on Computer Game,2012:39-44.
[5]T Kibret.A Bayesian Network Meta-analysis for Binary Outcome:A Simulation Study[J].Clinical Epidemiology(S1179-1349),2014,54(3):644-655.
[6]钟麟,佟明安,钟卫,等.影响图对策在多机协同空战的应用[J].北京航空航天大学学报,2007,33(4):450-453.ZHONG Lin,TONG Ming-an,ZHONG Wei,et al.Application of multistage influence diagram game theory for multiple cooperative air combat[J].Journal of Beijing University of Aeronautics and Astronautics,2007,33(4):450-453.
[7]钟麟.复杂条件下编队协同空战智能决策研究[D].西安:西北工业大学,2008:92-105.ZHONG Lin.Study on intelligent decision-making for cooperative team air combat under conditions of complexity[D].Xi’an:Northwestern ploytechnical university,2008:92-105.
[8]罗键,武鹤.基于交互式动态影响图的对手建模[J].控制与决策,2016,31(4):635-639.LUO Jian,WU He.Opponent modeling based on interactive dynamic influence diagram[J].Control and decision,2016,31(4):635-639.
[9]Julien Fatisson,Victor Oswald,Fran?ois Lalonde.Influence diagram of physiological and environmental factors affecting heart rate variability:an extended literature overview[J].Heart International(S1826-1868),2016,11(1):32-40.
[2]王广彦,贾希胜,李经中,等.遗传算法的经验公式仿真元模型的构建方法[J].火力与指挥控制,2011,36(10):119-123.WANG Guangyan,JIA Xisheng,LI Jingzhong,et al.The Study of Experiential Formula Simulation Metamodel Based on Genetic Algorithm[J].Fire Control&Command Control,2011,36(10):119-123.
[3]J P C Kleijnen.Kriging metamodeling in simulation:A review[J].European Journal of Operational Research(S0377-2217),2013,230(3):707-716.
[4]C S Longstreet.A meta-model for developing simulation games in higher education and professional development training[C].International Conference on Computer Game,2012:39-44.
[5]T Kibret.A Bayesian Network Meta-analysis for Binary Outcome:A Simulation Study[J].Clinical Epidemiology(S1179-1349),2014,54(3):644-655.
[6]钟麟,佟明安,钟卫,等.影响图对策在多机协同空战的应用[J].北京航空航天大学学报,2007,33(4):450-453.ZHONG Lin,TONG Ming-an,ZHONG Wei,et al.Application of multistage influence diagram game theory for multiple cooperative air combat[J].Journal of Beijing University of Aeronautics and Astronautics,2007,33(4):450-453.
[7]钟麟.复杂条件下编队协同空战智能决策研究[D].西安:西北工业大学,2008:92-105.ZHONG Lin.Study on intelligent decision-making for cooperative team air combat under conditions of complexity[D].Xi’an:Northwestern ploytechnical university,2008:92-105.
[8]罗键,武鹤.基于交互式动态影响图的对手建模[J].控制与决策,2016,31(4):635-639.LUO Jian,WU He.Opponent modeling based on interactive dynamic influence diagram[J].Control and decision,2016,31(4):635-639.
[9]Julien Fatisson,Victor Oswald,Fran?ois Lalonde.Influence diagram of physiological and environmental factors affecting heart rate variability:an extended literature overview[J].Heart International(S1826-1868),2016,11(1):32-40.