基于HLA的分布仿真系统数据采集解决方案

Solution of Data Collection in HLA

分布仿真系统运行数据采集是应用仿真系统的重要环节,各种采集策略都面临数据保存和恢复的问题。目前的基于数据库、文本文件和简单二进制文件的存储方法存在空间利用率低、速度慢、难以支持复杂数据类型、读取困难等缺点。针对集中式数据采集策略提出一种基于对象序列化的存取方案,通过将要记录的数据转换为对象,然后利用对象的序列化机制设计了一种新的二进制单一顺序数据文件的保存格式,并且结合采集器的动态生成及编译实现了其保存和恢复的接口自动生成。克服了现有的各种保存方式的弊端,为通用数据分析回放工具及采集数据的安全、加密压缩等提供了一个基础。

Data collecting is critical to simulation system based on HLA. Writing and reading of data collected is the most important part of DC. Currently the data is stored in database, text file or simple binary file. All the three methods need improved on the storage utilization, speed, support of complex data type and parsing of data stored. A new solution method based on object serialization is provided. The data that will be collected are expressed in object first, and stored in a binary file through object serialization. The result we get is a single binary ordinal file. Deserialization is the data reading and parsing method. A general data collecting tool using object serialization is available with compiler and code generation technology. This solution provides foundation for data playback, analysis, compression and encryption.