基于数据分发服务的通用仿真框架技术

General simulation framework based on data distribution service

分布式仿真技术应用场景的多样性使高层体系结构(HLA)的缺陷愈加明显:在数据传输规模较大时,实时性较低,无法应用于半实物仿真场景;HLA规范未对仿真应用的具体设计及规范进行定义,导致目前基于HLA的仿真系统在不同软硬件平台间兼容性差,降低了其可移植性及可重用性。数据分发服务(DDS)不仅能够覆盖HLA的核心特性,还能在保证仿真应用间的互操作性基础之上,进一步提高仿真系统的实时性、可重用性和移植性,因此,提出一套基于DDS的高实时性、高可靠性的仿真通用框架。该框架可为分布式仿真系统中各仿真应用的互操作提供通信支持,支撑并驱动仿真演练的执行。在框架中引入高精度时间同步技术,解决了DDS缺少时间管理机制的问题;针对框架提供了仿真开发套件,将底层通信、驱动逻辑透明化,减少仿真应用开发人员的工作量;提供了仿真监控工具,方便开发人员对仿真应用进行测试或执行训练。通过实践,验证了基于DDS的仿真框架技术可满足当下仿真对实时性要求高的需求。

High Level Architecture(HLA)is the mainstream technology of distributed simulation system. However,with the diversity of application scenarios,the drawbacks of HLA become more and more obvious. In the situation of largescale data transmission,the real-time performance of HLA is poor,and cannot be applied to hardware-in-the-loop simulation. HLA standards do not regulate the design of simulation applications,which results in the poor compatibility between different software and hardware platforms,reduces the portability and reusability of HLA-based simulation system.Data Distribution Service(DDS)can not only cover the core features of HLA,ensure the interoperability of simulation applications,but also improve the real-time,reusability and portability of simulation system. Therefore,a DDS-based general simulation framework of high real-time performance and reliability was proposed. The framework can provide communication support for the interoperability of simulation applications in distributed simulation system,and supports simulation execution. In order to solve the problem of lack of time management in DDS,high precision time synchronization technology was introduced,and a simulation developing kit was provided for the framework to transparentize the communication and driver logic and reduce the workload of developers. A simulation monitoring tool was provided for developers to conveniently test or train the simulation applications. It is verified that the DDS-based general simulation framework satisfies the high real-time requirements of current simulations.