基于多种群遗传算法的航天复杂系统测试任务调度  

Scheduling of aerospace complex system test tasks based on multi-population genetic algorithm

作　　者：胡涛[1] 申立群[1] 付晋 黄昌彬 HU Tao;SHEN Liqun;FU Jin;HUANG Changbin(School of Instrumentation Science and Engineering,Harbin Institute of Technology,Harbin 150001,China;Capital Aerospace Machinery Corporation Limited,Beijing 100076,China)

机构地区：[1]哈尔滨工业大学仪器科学与工程学院,哈尔滨黑龙江150001 [2]首都航天机械有限公司,北京100076

出　　处：《计算机集成制造系统》2024年第4期1255-1262,共8页Computer Integrated Manufacturing Systems

摘　　要：针对航天复杂系统型号较多,传统测试流程与调度设计只能人工定制化排布,效率较低且未有效优化,同时,考虑到航天复杂系统快速测试的迫切需求,提出一种基于多目标遗传算法的航天测试流程自动生成方法。该方法在测试项集合明确的前提下,将测试项抽象为离散事件,以测试总时间和测试资源均衡度为优化目标,充分考虑航天器测试的诸多约束,将其作为遗传算法执行过程中交叉或变异的禁忌项。在初始种群确定后,对测试流程和调度方案进行自动生成和优化。对算例的仿真结果表明,该方法相对于同实验条件下的传统半串行测试方法和单目标优化方法,测试总时间或资源均衡度得到了较大提升。在进一步扩展优化目标和约束项后,该方法可有效提高航天复杂系统测试过程的快速响应能力和可靠性。ed as discrete events,the total test time and test resource balance were taken as the optimization goals,and many constraints of spacecraft testing were fully considered as the taboo items in the execution process of the genetic algorithm or variant contraindications.After the initial population was determined,the aerospace testing process was automatically generated and optimized.The simulation results of the numerical example showed that compared with the traditional semi-serial test method and single-objective optimization method under the same experimental conditions,the total test time or resource balance had been greatly improved.After further expanding the optimization objectives and constraints,the method could effectively improve the rapid response capability and reliability of the aerospace complex system testing process.

关 键 词：流程优化 多种群遗传算法 并行任务调度 航天复杂系统测试 

分 类 号：TJ765.4[兵器科学与技术—武器系统与运用工程] TP18[自动化与计算机技术—控制理论与控制工程]