摘要： 在现有仿真时间推进机制中，事件的发生时间采用精确的时间戳，但是有时事件的发生时间用一个时间区间来表达更合理。该文提供了一种新的时间卷曲方法来发掘事件发生时间的不确定性，此方法的核心思想是：用事件的时间区间模型代替精确时间戳模型来提高事件并发性；当事件的时间区间重叠时用因果关系防止失序；对因果关系事件采用语义学方法进行区分，防止过多回滚。在RTI上增加时间管理扩展组件，用来实现基于时间区间的乐观时间同步机制。

关键词: HLA, 时间卷曲, 时间戳, 时间不确定性

Abstract: Current works in distributed discrete event simulations depend on precise time stamps to events, but sometimes it is reasonable for modelers to use time intervals. This paper describes a time warp mechanism to exploit temporal uncertainty. The basic idea of the approach is to provide an event model in which events are assigned time intervals to exploit parallelism, a sub-mechanism to preserve causal ordering when time intervals to events overlap and a semantics sub-mechanism to prevent rollback too much. A component based on RTI is implemented, which ensures a practical time warp mechanism.

Key words: HLA, Time warp, Time stamp, Temporal uncertainty

中图分类号: 

• TP338