Multi-disciplinary joint simulation method and system under distributed architecture

Abstract

The invention discloses a multi-disciplinary joint simulation method under a distributed architecture, which realizes model parallel computing and data parallel communication in the distributed architecture, can realize distributed computing solution of a simulation model, and adopts parallel communication at a server side in the architecture; and secondary improvement of simulation efficiency is achieved through performance statistics, calculation time consumption of a model corresponding to each port is counted in real time in the simulation process based on data parallel communication of a server side, and the simulation efficiency of all models is printed and displayed after simulation is finished. Under the same simulation model condition, the distributed architecture integrated simulation can improve the calculation efficiency by about one order of magnitude. The larger the scale of the models participating in integration is, the more obvious the efficiency improvement is; and a brand new solution is provided, so that simple and convenient integration of heterogeneous models becomes possible.

Description

technical field [0001] The invention belongs to the technical field of simulation, and relates to a multidisciplinary joint simulation method and system under a distributed framework. Background technique [0002] The essence of system simulation is to use the numerical integral algorithm to numerically solve the differential initial value problem. Constrained by the solution algorithm and solution mechanism, no matter what the size of the simulation model, it can only call a single thread to solve it. When the size of the simulation model increases to a certain extent (the size of the model mainly depends on the number of variables and the rigidity of the equation system, because it contains multiple disciplines and the number of sub-models is large, the model of the general virtual performance prototype or the model integration simulation performed by the host Generally, the scale is very large), it is difficult to efficiently complete the simulation task through a solver....

AI Technical Summary

Problems solved by technology

However, the defects of these two methods are obvious: only the co-simulation format fmu export can integrate the model built in it, and there can be no error reporting, which requires a high technical level of the modeler to achieve this; and through FMI The official website query shows that there are very few simulation tools that can support fmu export in the Co-Simulation format. Many mainstream simulation tools are more willing to import the fmu model of other tools instead of exporting their own models to other tools due to commercial interests and intellectual property considerations. ;

[0006] To sum up, although there have been different solutions for distributed integrated simulation at home and abroad, so far there is no public information that clearly points out how to carry out the work if you want to further improve the simulation efficiency after model integration

Images