针对传统的规则学习算法很难解决顶层决策，以及现阶段航空集群作为一种新兴的作战样式，没有太多现成的数据和案例可供参考，暂不具备"从战争中学习战争"条件等难题，从战争设计的角度出发，探讨性地提出了一种基于情景分析的规则库构建方法。首先，从系统演化的外部触发条件和内部驱动机制出发，提出了基于事件触发-规则驱动的自主决策机制；然后，将航空集群决策规则拆分为事件、条件、动作3部分，并采用事件-条件-动作（ECA）描述机制进行规范化表达；最后，借鉴情景分析理论的思想，通过详细分析作战过程的逻辑关系、状态变迁，实现对象、事件、行为的关联，并以无人机自主察打任务规则提取为例进行了验证分析。
The traditional algorithm of rule learning is difficult to solve the problem of top-level decision-making. In addition, as a new type of combat, aviation swarm does not have sufficient labeled data and battle cases for reference at this stage, thus the condition for ‘learn to fight from the war’ is temporarily immature. To solve the above problems, this paper proposes a construction method of rule base based on the scenario analysis from the perspective of war design. Firstly, based on the external triggering conditions and internal driving mechanism of system evolution, an event-triggered and rule-driven autonomous decision-making mechanism is proposed. Secondly, the decision rules of aviation swarm are divided into event, condition and action and are expressed by the Event-Condition-Action (ECA) description mechanism. Finally, by adopting the idea of scenario analysis theory, the logic and state transition of the operational process are analyzed in detail, and the associations among objects, events, and behaviors are finally realized. Moreover, the rules extraction of autonomous reconnaissance and strike mission using UAV swarm is carried out to verify the proposed theory.
[1] 柏鹏, 梁晓龙, 王鹏, 等. 新型航空集群空中作战体系研究[J]. 空军工程大学学报(军事科学版), 2016, 16(2):1-4. BAI P, LIANG X L, WANG P, et al. Research on new air combat system of aviation swarms[J]. Journal of Air Force Engineering University (Military Science Edition), 2016, 16(2):1-4(in Chinese).
[2] FORTUNY E J D, MARTENS D. Active learning-based pedagogical rule extraction[J]. IEEE Transactions on Neural Networks and Learning Systems, 2015, 26(11):2664-2677.
[3] POURPANAH F, LIM C P, SALEH J M. A hybrid model of fuzzy ARTMAP and genetic algorithm for data classification and rule extraction[J]. Expert Systems With Applications, 2016, 49:74-85.
[4] MASHAYEKHI M, GRAS R. Rule extraction from decision trees ensembles:New algorithms based on heuristic search and sparse group lasso methods[J]. International Journal of Information Technology & Decision Making, 2017, 16(6):1707-1727.
[5] KARIMI T, FORREST J Y L. Using grey incidence to analyze the energy audit reports and rough set for rule extraction[J]. Kybernetes, 2016, 45(7):1024-1035.
[6] YANG H F, ZHANG J F, HU L H. Classification rule extraction based on the rough concept lattice[J]. Kybernetes, 2010, 39(8):1336-1343.
[7] LI A, CHEN G. A new approach for rule extraction of expert system based on SVM[J]. Measurement, 2014, 47:715-723.
[8] SHAHABI M, HASSANPOUR H, MASHAYEKHI H. Rule extraction for fatty liver detection using neural networks[J]. Neural Computing & Applications, 2017, 31(4):979-989.
[9] 刘晓伶, 卢涛. 基于情景分析的ECA规则提取方法研究[J]. 计算机工程, 2012, 38(22):154-158. LIU X L, LU T. Research on ECA rule extraction method based on scenario analysis[J]. Computer Engineering, 2012, 38(22):154-158(in Chinese).
[10] PISHVAEE M S, FATHI M, JOLAI F. A fuzzy clustering-based method for scenario analysis in strategic planning:The case of an Asian pharmaceutical company[J]. South African Journal of Business Management, 2008, 39(3):21-31.
[11] SANGHA K K, GERRITSEN R, JEREMY R S. Repurposing government expenditure for enhancing Indigenous well-being in Australia:A scenario analysis for a new paradigm[J]. Economic Analysis and Policy, 2019, 63:75-91.
[12] WRIGHT G, CAIRNS G, BRIEN F A, et al. Scenario analysis to support decision making in addressing wicked problems:Pitfalls and potential[J]. European Journal of Operational Research, 2019, 278(1):3-19.
[13] CLEMONS E K. Scenario analysis and managing strategic ambiguity:How to remember future events, before they actually occur![M]. Heidelberg:Springer, 2019:163-165.
[14] 梁晓龙, 何吕龙, 张佳强, 等. 航空集群构型控制及其演化方法[J]. 中国科学:技术科学, 2019, 49(3):277-287. LIANG X L, HE L L, ZHANG J Q, et al. Configuration control and evolutionary mechanism of aircraft swarm[J]. Scientia Sinica (Technologica), 2019, 49(3):277-287(in Chinese).
[15] ISAZADEH A, PEDRYCZ W, MAHAN F. ECA rule learning in dynamic environments[J]. Expert Systems with Applications, 2014, 41(17):7847-7857.
[16] 毛琼, 李小民, 王正军. 基于规则的无人机编队队形构建与重构控制方法[J]. 系统工程与电子技术, 2019, 41(5):1118-1126. MAO Q, LI X M, WANG Z J. Formation and reformation control method for UAVs formation shape based on rules[J]. Systems Engineering and Electronics, 2019, 41(5):1118-1126(in Chinese).
[17] REYNOLDS C W. Flocks, herds and schools:A distributed behavioral model[J]. Computer Graphics, 1987, 21(4):25-34.
[18] MCCARTHY D R, DAYAL U. The architecture of an active database management system[C]//Proceedings of the 1989 ACM SIGMOD International Conference on Management of Data. New York:ACM, 1989:215-224.
[19] ZHOU Y L, HSU M. A theory for rule triggering systems[C]//International Conference on Extending Database Technology. Heidelberg:Springer, 1990:409-421.
[20] 金欣. 指挥控制智能化现状与发展[J]. 指挥信息系统与技术, 2017, 8(4):10-18. JIN X. Status and development of intelligent command and control[J]. Command Information System and Technology, 2017, 8(4):10-18.
[21] SHI Z, WEI C A, FU P, et al. A parallel search strategy based on sparse representation for infrared target tracking[J]. Algorithms, 2015, 8(3):529-540.
CJA