系统工程与电子技术 ›› 2022, Vol. 44 ›› Issue (7): 2201-2210.doi: 10.12305/j.issn.1001-506X.2022.07.16

• 系统工程 • 上一篇    下一篇

任务主体二元约束下作战任务分解EVA方法

刘乾*, 鲁云军, 陈克斌, 韩梦瑶, 郭亮   

  1. 国防科技大学信息通信学院, 湖北 武汉 430010
  • 收稿日期:2021-05-28 出版日期:2022-06-22 发布日期:2022-06-28
  • 通讯作者: 刘乾
  • 作者简介:刘乾(1989—), 男, 博士研究生, 主要研究方向为任务规划、作战仿真|鲁云军(1973—), 男, 教授, 博士, 主要研究方向为军事运筹、指挥信息系统建模与仿真|陈克斌(1987—), 男, 博士研究生, 主要研究方向为指挥信息系统体系建模、体系能力研究|韩梦瑶(1989—), 女, 博士研究生, 主要研究方向为复杂网络建模、因果网络研究|郭亮(1985—), 男, 博士研究生, 主要研究方向为军事运筹、网络信息体系建模与评估
  • 基金资助:
    国家社会科学基金(2020-SK-C-104);全军军事类研究生资助课题重点项目(JY2019B071)

Combat task decomposition EVA method based on binary constraints of task subject

Qian LIU*, Yunjun LU, Kebin CHEN, Mengyao HAN, Liang GUO   

  1. College of Information and Communication, National University of Defense Technology, Wuhan 430010, China
  • Received:2021-05-28 Online:2022-06-22 Published:2022-06-28
  • Contact: Qian LIU

摘要:

针对复杂作战任务分解中存在的随意性、不确定性问题, 综合考虑任务主体能力属性和结构特征等二元约束, 提出了一种由子任务集提取(extraction, E)、约束检验(verification, V)、子任务集调整(adjustment, A)等步骤递进循环形成的任务分解EVA方法。首先, 构建了全局任务空间, 提出基于任务匹配的子任务集提取方法; 其次, 针对任务主体能力属性和结构特征的二元约束, 建立了子任务集调整模型, 通过改进精英保留策略, 引入任务分解粒度和交叉变异概率动态调整策略, 提出了一种引进的非支配排序遗传算法-Ⅱ(improved non-dominated sorting genetic algorithm-Ⅱ, INSGA-Ⅱ)算法; 最后, 进行仿真实验, 验证了算法相较于传统多目标优化算法在解集多样性、收敛性和时间性能上的优势。研究结果表明, 所提方法能够使决策者依据任务主体实际自主调控任务分解结果, 在一定程度上克服了传统方法过度依赖主观经验, 忽略任务主体能力属性、结构特征约束的问题。

关键词: 作战任务分解, 任务主体, 二元约束, 任务提取-约束检验-子任务集调整方法, 多目标优化, 任务分解

Abstract:

Aiming at the arbitrariness and uncertainty in the decomposition of complex combat tasks, a extraction-verification-adjustment (EVA) task decomposition method is proposed, which is formed by the extraction, verification, and adjustment of the sub task set, considering the binary constraints such as the capability attributes and structural characteristics of the task subject. First, the global task space is constructed, and the subtask set extraction method based on task matching is proposed. Secondly, according to the binary constraints of the task subject's ability attributes and structural characteristics, a quantitative adjustment model of subtask sets is established. By improving the elitist retention strategy and introducing task decomposition granularity and the dynamic adjustment strategy of crossover mutation probability, the improved non-dominated sorting genetic algorithm-Ⅱ (INSGA-Ⅱ) algorithm is proposed. Finally, the simulation results verify that INSGA-Ⅱ has the advantages of diversity, convergence in solution set and timeliness performance compared with the traditional optimization algorithms. The results show that the method proposed in this paper enable decision makers to adjust and control the task decomposition results according to the actual situation of the task subject, and overcome the problem that the traditional methods rely on subjective experience and ignore the constraints of the ability attributes and structural characteristics of the task subject.

Key words: combat task decomposition, task subject, binary constraints, extraction-verification-adjustment (EVA) method, multi-objective optimization, task descomposition

中图分类号: