测试系统多智能体建模技术及应用

Modeling and application of multi-agents oriented to test systems

现代测试系统需要管理多种测试资源,并利用它们之间的高效自主协调完成测试任务,但是系统软件的设计却缺乏统一模型框架和理论分析依据。将智能体技术应用于测试系统,建立了基于多智能体的测试系统结构模型和逻辑模型,然后利用Petri网建模方法对多智能体测试系统的运行过程进行形式化建模,并通过可达图分析,验证了该模型的可达性、安全性与活性,最后针对固体继电器的性能参数测试开发实例系统,在工程应用中验证了该智能建模技术的有效性,结果证明实例系统的运行过程与模型仿真结果一致,无冲突和死锁等现象,并与未采用智能体技术开发的测试系统作对比实验,表明当测试任务或测试资源增多时,智能测试系统的优越性更为明显。

The management and cooperation of multiple test resources are important in modem test systems for completing test tasks correctly and efficiently, but the system software design lacks unified model framework and theoretical analysis basis. In order to improve the intelligence and efficiency of test systems, the architecture and logical model of test systems based on multi-agents are built. Then the formal model is built for the running process of the test system with Petri net, and the reaehability, safety as well as activity of the model are verified. In order to prove the validity of this method and modeling technology, an example system for the performance parameter testing of solid state relays was build. The running process of the example system is deadlock-free and conflict-free, which is consistent with the simulation results of the Petri net model. Compared with the test system developed without using agent technique, test result shows that the test system based on multi-azents is more effective while the test tasks and resources increase.