信息处理单元测试系统并行测试技术研究

Research on Parallel Test Technology of Information Processing Unit Test System

摘要:

针对信息处理单元测试的功能需求,设计实现了基于PXIe总线的测试系统,系统包括测控单元、供电单元和接口单元三个组成部分,并完成了测试系统的软件设计；根据信息处理单元测试系统实际的测试流程和所需资源的使用情况,将并行测试技术引入到系统中,重新对相互独立的测试任务进行了并行化分析,构建了该测试系统的任务调度数学模型,提出了一种改进的自适应遗传算法(即IAGA算法),解决了并行测试模块中复杂且难以优化的任务调度问题；对任务调度模型进行了算法仿真验证和实验结果分析,验证了所得解的全局性,所得出的执行序列与串行测试相比,测试效率提高了43.57%,并与其他算法进行了对比,验证了IAGA算法的可行性与优越性；最后将IAGA算法嵌入到了测试系统的软件当中,实现了该优化策略的实际工程应用。

Abstract:

According to the functional requirements of information processing unit test, a test system based on PXIe bus was designed and implemented. The system was composed of three parts: measurement and control unit, power supply unit and interface unit, and the software design of the test system was completed. According to the actual test process and the use of required resources of the information processing unit test system, the parallel test technology is introduced into the system, and the parallel analysis of the independent test tasks is carried out again. The task scheduling mathematical model of the test system is constructed, and an improved adaptive genetic algorithm (IAGA algorithm) is proposed. The complicated task scheduling problem which is difficult to optimize in parallel test module is solved. The algorithm simulation verification and experimental results analysis of the task scheduling model are carried out to verify the global solution. Compared with the serial test, the test efficiency of the obtained execution sequence is increased by 43.57%. Compared with other algorithms, the feasibility and superiority of the IAGA algorithm are verified. Finally, the IAGA algorithm is embedded into the software of the test system to realize the practical engineering application of the optimization strategy.