【摘要】 研究社会灾害和生态环境中的发慌问题,需科学决策。针对复杂系统的研究涉及到多种学科交叉,进行多Agent建模仿真需要面对不同领域知识融合的问题。在最新多Agent建模仿真平台Repast Simphony(简称Repast S)的研究基础上,为决策者提供依据,提出以其层次建模框架满足多专家群体协作建模需求,以Score元模型为建模基础进行领域知识融合的策略。以森林火灾系统为例,应用Repast S进行了建模仿真实践。所建立模型有机融合了林火蔓延的元胞自动机模式和扑救协作的社会网络模式等领域知识,进行仿真。仿真结果体现出各种模式对森林火灾系统演化态势的影响。Repast S为复杂系统学科交叉的多Agent建模仿真提供了有效支持。更多还原

【Abstract】 Research on complex systems sometimes involves intersections between disciplines,and so combination of knowledge from different fields should be considered in Agent-based modeling and simulation for them.According to Repast Simphony,a new advanced Agent-based modeling and simulation toolkits,group hierarchical modeling structure is applied to promote collaborative modeling for experts,and Score meta-model is implemented to facilitate combination of knowledge.As an example,a forest fire system model is developed using Repast Simphony.The model integrates knowledge from different fields such as cellular automata pattern for fire spreading and network coordination pattern for fire fighting.The simulation results show the influence of these patterns towards the development tendency of the forest fire system.Repast Simphony can effectively support Agent-based modeling and simulation for complex systems with intersections between disciplines.更多还原