协同仿真算法研究综述

A Review of Co-simulation Algorithm

协同仿真算法近年来广泛用于数字孪生-多物理域耦合分析以及复杂大规模动力学系统并行计算。相比传统强耦合分析,具有模块化、智能化和非侵入高效计算等优点。随着智能制造发展和工业互联网深度融合,基于协同仿真算法的数据接口和计算平台迅速发展。由于多物理域模型刚度、精度需求和数值特性的差异,协同仿真算法数值稳定性一直是制约其应用的瓶颈问题。首先,介绍了协同仿真算法的基本概念、分类和耦合模型。然后,分别讨论了界面变量插值引入的数值稳定性问题,并汇总了协同仿真增稳技术、局部误差分析和宏观时间步长控制方法。其次,总结了协同仿真算法在综合能源系统、动力学系统和数字孪生技术中的应用和算法平台。最后,面向未来高效高稳定性协同仿真算法,对协同仿真算法研究和应用方面进行了总结和展望。

The co-simulation algorithm has been widely used in digital twin,multi-physics domain coupling analysis and parallel computing of complex large-scale dynamic systems.Compared with traditional strongly coupled analysis,it has the advantages of modularity,intelligence,non-intrusive and efficient calculation.With the development of intelligent manufacturing and the deep integration of the industrial Internet,data interfaces and computing platforms based on co-simulation algorithms have developed rapidly.Due to the differences in stiffness,accuracy requirements and numerical characteristics of multi-physics domain models,the numerical stability of co-simulation algorithms has always been a bottleneck restricting their application.First,the basic concepts,classification and coupling models of co-simulation algorithms are introduced.Then,the numerical stability issues introduced by interface variable interpolation are discussed respectively,and the co-simulation stabilization techniques,local error analysis and macro-time step control methods are summarized.Next,the application and algorithm platform of co-simulation algorithms in integrated energy systems,dynamic systems and digital twin technology are summarized.Finally,facing the future of efficient and high-stability co-simulation algorithms,the research and application of co-simulation algorithms are summarized and prospected.