欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了保证面向指挥决策支持的在线仿真系统可信性,提出了情报数据驱动的在线仿真系统动态修正方法。首先,依据情报数据变化快慢,将其分为缓变和快变两类数据;然后,从位置接近性和外形相似性两个方面刻画缓变数据间差异,从趋势项差异和平稳项差异两个方面刻画快变数据间差异,由此实现了仿真输出与情报数据一致性的度量;进一步,基于拉丁超立方实验设计方法确定仿真模型＂修正集合＂,并利用主成分分析综合仿真输出一致性指标,从修正集合中选择使得仿真模型可信性最佳的修正方案。通过应用实例,表明了方法的有效性。
英文摘要：
To ensure the creditability of the online simulation system for decision-making, the method for dynamically modifying the online simulation based on intelligence data-driven is proposed. The intelligence data are divided into gradual data and fast data according to the change ratio. For describing the consistency between the simulation output and the intelligence data, the differences between the gradual data are depicted by the proximity of the position and the similarity of the shape, the fast data is decomposed into the trend item and sta- tionary item, and the measure models of differences for each item are given. Furthermore, the modification set of the simulation model is gained via Latin hypercube sampling, the best modification scheme is selected by in- tegrating the consistency indexes of the simulation output based on principal component analysis. Finally, the validity of the method is shown in the application.
同期刊论文项目
同项目期刊论文
期刊信息
位置：