不确定环境下作战体系结构适应性优化研究

Operational system-of-systems structure adaptive optimization problem is one of the core issues in command and control domain. It is also the key point to solve the problem of rapid reconfiguration for operation resources and to obtain system-of-systems antagonism’s advantage in complex and changeful battle field. In this project, three kinds of theory, which include supernetwork theory, receding-horizon control theory and quantum evolution theory, will be used to focus on the study of the dynamic delaminating receding optimizing control method for operational system-of-systems structure oriented to dynamic uncertain environment. Firstly, based on the constructed operational system-of-systems supernetwork description model and system-of-systems capability measure model, the influence of the structure ability measure when supernetwork topology varies is analysed and the interrelationship between structure variables and capability measure in system-of-systems structure is revealed, and the delaminating optimizing control method for the operational system-of-systems supernetwork is proposed. Secondly, the receding-horizon policy is introduced to divide the full-time domain system-of-systems structure adaptive optimizing control problem into several times hierarchical loop repetition short-time domain optimization problem, then the computing approach of receding-horizon cycle based on situation prediction is proposed. Finally, the sub-problems network optimization models in receding-horizon cycle are built, which can be solved by the presented algorithms including chaos quantum immune clone algorithm and two-layer quantum particle swarm optimization algorithm, the algorithms are all relative to the network characteristics of sub-problems and have neighborhood search mechanism. The research achievements will not only be helpful to improve the system-of-systems structure adaptive optimization theory, but also provide advanced theory guide and technical support for future’s quick combat command and control.

作战体系结构适应性优化问题是指挥控制领域最核心的问题之一，是解决复杂多变战场空间中作战资源快速重组，并获取体系对抗优势的关键。本项目拟将超网络理论、时域滚动控制理论、量子进化理论相结合，研究面向不确定环境下的作战体系结构动态分层滚动优化控制方法。建立作战体系超网络描述模型、体系能力测度模型，分析超网络拓扑结构变化对于体系能力测度的影响，揭示体系结构中结构变量与能力测度的内在联系，提出作战体系超网络的分层优化控制方法；设计时域滚动策略，将全时域体系结构适应性优化控制分解为递阶循环的多次短时域优化，提出基于态势预测的时域滚动周期确定方法；构建时域滚动周期内子问题网络优化控制模型，提出与各子问题网络特点相适应的、具有邻域搜索机制的混沌量子免疫克隆算法、量子粒子群算法等进行优化模型的快速求解。预期成果将进一步完善体系结构适应性优化理论，为未来实施敏捷的战场指挥控制提供新的理论指导和技术途径。

本项目针对作战体系具有动态不确定性以及复杂度高等特点，引入超网络理论，研究基于超网络的作战体系建模描述方法，分析超网络拓扑结构变化对于体系能力测度的影响，揭示体系结构中结构变量与能力测度的内在联系，提出作战体系超网络的分层优化控制方法；针对任务-平台资源的网络的优化问题，建立了以任务总完成时间最小、任务执行质量最高、协同效能最大的多目标优化任务分配模型，提出基于动态列表规划和多目标烟花算法的模型求解方法。针对指挥协同结构优化控制问题，提出了以作战实体行动达成一致为目标的网络结构设计方法，建立了网络拓扑有向图及其等效图，提出了协同一致性协议及协议收敛性的相关定理及其证明。针对信息网络结构适应性优化控制问题，提出了基于信息价值需求的反导体系结构超网络均衡优化方法。其研究成果不仅可以提升作战体系的战场空间适应能力，降低作战体系结构全时域建模和求解的复杂度，还能够提升作战体系中实体的自组织和自协作能力，进而提高体系的整体作战效能。取得的研究成果主要反映在：发表期刊学术论文36篇，会议论文5篇，其中SCI检索论文14篇，EI检索11篇。出版专著3部。获得国防发明专利授权5项，申请受理国防发明专利2项。获得空军防空反导优秀研究成果一等奖2项、二等奖1项、三等奖励1项。