A Cloud Simulation Based Environment for Multi-disciplinary Collaborative Simulation and Optimization

Liqin Guo,Mei Wang,Chao Ruan,Ting Yu Lin,Chen Yang,Lichao Wei,Chao Geng,Chi Xing,Yingying Xiao

DOI: https://doi.org/10.1109/cscwd.2017.8066735

2017-01-01

Abstract:Multi-disciplinary virtual prototype (MDVP) based on modeling and computer simulation technology has been applied in a wide range of engineering applications, especially the design, testing and evaluation of complex products. Generally, a large set of parameters need to be optimized to improve the performance of the prototype, which comprises a number of heterogeneous models from multiple domains. This calls for numerous collaborative simulation experiments in heterogeneous computing environments, which are difficult to build and configure. Cloud simulation is a promising solution with the extremely large resource pool and dynamic construction of virtual computing environments. Thus this paper firstly proposed an optimization framework of MDVP which formulated the domain models, optimizer and the distributed interactive environment on cloud for multi-disciplinary engineers. Then we designed the procedure for conducting parallel optimization of MDVP in the cloud simulation system by con-current execution of MDVP simulation system. The method proposed in the paper has been applied to a MDVP in the aerospace industry, which indicates that the proposed method can support an efficient engineering methodology that can transform the traditional centralized, serial simulation optimization to the distributed collaborative and parallel simulation optimization.