动态数据驱动仿真综述
Dynamic Data Driven Simulation:An Overview
谢旭 1邱晓刚 1包亦正 1许凯1
作者信息
- 1. 国防科技大学系统工程学院,湖南长沙 410073
- 折叠
摘要
动态数据驱动仿真是一种"模型和数据相结合"的仿真范式,它将真实系统的观测(数据)持续注入仿真(模型),让数据动态地校正仿真(状态、参数),以此来提高基于仿真的估计和预测能力.由于动态数据驱动仿真融合了模型预测和实时观测两方面的信息,因此它能更准确地估计系统状态并预测状态的未来演化.梳理了动态数据驱动仿真的思想起源和基本概念,延伸介绍了"模型和数据相结合"的思想孕育的一系列仿真范式,并辨析了它们之间的联系和区别;介绍了基于粒子滤波的数据同化方法和identical-twin仿真实验方法;从应用场景、模型和数据、数据同化算法、与新技术的融合等4个维度综述了动态数据驱动仿真的研究现状;从仿真模型、观测数据、数据同化、运行效率、应用领域等5个方面对动态数据驱动仿真未来研究方向进行了展望.
Abstract
Dynamic data driven simulation is a simulation paradigm which integrates simulation and data together.This paradigm continuously feeds real-time data into the simulation,enabling the simulation be dynamically adjusted by the data,which thus improves the simulation-based estimation and prediction capability.Due to this integration,the dynamic data driven simulation can estimate system states and predict future state evolution more accurately.This paper reviews the origins and basic concept of dynamic data driven simulation,and introduces several simulation paradigms originated from the idea of"integrating models with data",and identifies the linkages and differences among them.The particle filter-based data assimilation method and the identical-twin simulation experiment are introduced.The current research status of dynamic data driven simulation is summarized from four perspectives,i.e.,application scenarios,models and data,data assimilation algorithms,and integration with new technologies.Finally,the future research directions are outlooked from five aspects,which are simulation models,measurement data,data assimilation,algorithm performance,and application areas.
关键词
建模与仿真/动态数据驱动仿真/数据同化/粒子滤波/identical-twin仿真实验Key words
modeling and simulation/dynamic data driven simulation/data assimilation/particle filters/identical-twin simulation experiment引用本文复制引用
出版年
2024
NETL
NSTL
万方数据