基于RF-XGBoost算法的无人机多回合攻防博弈决策

doi:10.12305/j.issn.1001-506X.2025.02.18

摘要/Abstract

摘要：

为解决不平衡空战数据集下的无人机多回合博弈对抗问题, 提出一种随机森林-极限梯度提升(random forest-eXtreme gradient boosting, RF-XGBoost)算法以进行攻防博弈决策研究。通过分析红蓝双方的运动状态和空战信息, 建立支付矩阵模型, 利用线性归纳法求解当前博弈纳什均衡解和期望收益, 以蓝方最终获胜作为博弈对抗是否停止的判断条件。在博弈对抗过程中, 首先基于随机森林(random forest, RF)算法对空战数据集进行特征降维以提高空战决策的实时性, 然后提出改进的XGBoost算法来处理不平衡数据集, 将其用于确定最优机动动作以提高机动决策准确率和提升蓝方对抗态势, 并得到下一回合的红蓝空战信息; 之后, 根据下一回合的支付矩阵模型重新计算纳什均衡解和期望收益, 直至蓝方获胜; 最后, 通过仿真验证所提算法的可行性和有效性。

关键词: 无人机, 随机森林, 极限梯度提升, 多回合博弈

Abstract:

To solve the multi-round game confrontation problem of unmanned aerial vehicles (UAVs) with unbalanced air combat data set, a random forest-eXtreme gradient boosting (RF-XGBoost) algorithm is proposed to study the attack and defense game decision-making. The payment matrix model is established by analyzing the movement status and air combat information of the red and blue sides, then, the linear induction method is considered to solve the current Nash equilibrium solution and expected return of the game, and whether the game confrontation will stop depends on the victory of the blue side. In the process of game confrontation, in the first place, the feature dimensionality reduction of air combat data set is conducted based on the random forest (RF) algorithm to improve the real-time performance of air combat decision-making. Then, an improved XGBoost algorithm is proposed to deal with the unbalanced data set, which is used to determine the optimal maneuvers to improve the accuracy of maneuver decision-making and enhance blue confrontation's situation, and air combat information of the next round of red and blue sides is obtained. Furthermore, the Nash equilibrium solution and expected return based on the payment matrix model of the next round can be obtained once again, until the blue side wins. Finally, the feasibility and effectiveness of the proposed algorithm are verified by simulation test.

Key words: unmanned aerial vehicle (UAV), random forest (RF), eXtreme gradient boosting (XGBoost), multi-round game

中图分类号:

TP391.9

邹世培, 王玉惠, 刘鸿睿. 基于RF-XGBoost算法的无人机多回合攻防博弈决策[J]. 系统工程与电子技术, 2025, 47(2): 518-526.

Shipei ZOU, Yuhui WANG, Hongrui LIU. Multi-round attack and defense game decision-making of UAVs based on RF-XGBoost algorithm[J]. Systems Engineering and Electronics, 2025, 47(2): 518-526.

图/表 14

图1

表1

表2

图2

表3

表4

表5

图3

图4

表6

图5

图6

图7

图8

参考文献 30

1	LI B , HUANG J Y , BAI S X , et al. Autonomous air combat decision-making of UAV based on parallel self-play reinforcement learning[J]. CAAI Transactions on Intelligence Technology, 2022, 8 (1): 64- 81.
2	DONG Y Q , AI J L , LIU J Q . Guidance and control for own aircraft in the autonomous air combat: a historical review and future prospects[J]. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2019, 233 (16): 5943- 5991. doi: 10.1177/0954410019889447
3	ZHANG J D , YU Y F , ZHENG L H , et al. Situational continuity-based air combat autonomous maneuvering decision-making[J]. Defence Technology, 2023, 29 (11): 66- 79.
4	HE Y X , DU Z J , HUANG L W , et al. Maneuver decision-making method for ship collision avoidance in Chengshantou traffic separation scheme waters[J]. Applied Sciences, 2023, 13 (14): 8437. doi: 10.3390/app13148437
5	LU H C , WU B Y , CHEN J Q . Fighter equipment contribution evaluation based on maneuver decision[J]. IEEE Access, 2021, 9, 132241- 132254. doi: 10.1109/ACCESS.2021.3115142
6	董一群, 艾剑良. 自主空战技术中的机动决策: 进展与展望[J]. 航空学报, 2020, 41 (S2): 4- 12.
	DONG Y Q , AI J L . Decision making in autonomous air combat: review and prospects[J]. Acta Aeronautica et Astronautica Sinica, 2020, 41 (S2): 4- 12.
7	AUSTIN F , CARBONE G , HINZ H , et al. Game theory for automated maneuvering during air-to-air combat[J]. Journal of Guidance, Control, and Dynamics, 1990, 13 (6): 1143- 1149. doi: 10.2514/3.20590
8	ARDEMA A , HEYMANN M , RAJAN N . Analysis of a combat problem: the turret game[J]. Journal of Optimization Theory and Applications, 1987, 54 (1): 23- 42. doi: 10.1007/BF00940403
9	WONG R E . Some aerospace differential games[J]. Journal of Spacecraft and Rockets, 2012, 4 (11): 1460- 1465.
10	VIRTANEN K , KARELAHTI J , RAIVIO T . Modeling air combat by a moving horizon influence diagram game[J]. Journal of Guidance, Control, and Dynamics, 2012, 29 (5): 1080- 1091.
11	谭目来, 丁达理, 谢磊, 等. 基于模糊专家系统与IDE算法的UCAV逃逸机动决策[J]. 系统工程与电子技术, 2022, 44 (6): 1984- 1993. doi: 10.12305/j.issn.1001-506X.2022.06.26
	TAN M L , DING D L , XIE L , et al. UCAV escape maneuvering decision based on fuzzy expert system and IDE algorithm[J]. Systems Engineering and Electronics, 2022, 44 (6): 1984- 1993. doi: 10.12305/j.issn.1001-506X.2022.06.26
12	傅莉, 谢福怀, 孟光磊, 等. 基于滚动时域的无人机空战决策专家系统[J]. 北京航空航天大学学报, 2015, 41 (11): 1994- 1999.
	FU L , XIE F H , MENG G L , et al. An UAV air-combat decision expert system based on receding horizon control[J]. Journal of Beijing University of Aeronautics and Astronautics, 2015, 41 (11): 1994- 1999.
13	ZHANG H P , ZHOU H , WEI Y J , et al. Autonomous maneuver decision-making method based on reinforcement learning and Monte Carlo tree search[J]. Frontiers in Neurorobotics, 2022, 16, 996412. doi: 10.3389/fnbot.2022.996412
14	何旭, 景小宁, 冯超. 基于蒙特卡罗树搜索方法的空战机动决策[J]. 空军工程大学学报(自然科学版), 2017, 18 (5): 36- 41.
	HE X , JING X N , FENG C . Air combat maneuver decision based on MCTS method[J]. Journal of Air Force Engineering University(Natural Science Edition), 2017, 18 (5): 36- 41.
15	TENG T H, TAN A H, TAN Y S, et al. Self-organizing neural networks for learning air combat maneuvers[C]//Proc. of the International Joint Conference on Neural Networks, 2012.
16	LI B , LIANG S Y , CHEN D Q , et al. A decision-making method for air combat maneuver based on hybrid deep learning network[J]. Chinese Journal of Electronics, 2022, 31 (1): 107- 115.
17	张宏鹏, 黄长强, 轩永波, 等. 基于深度神经网络的无人作战飞机自主空战机动决策[J]. 兵工学报, 2020, 41 (8): 1613- 1622. doi: 10.3969/j.issn.1000-1093.2020.08.016
	ZHANG H P , HUANG C Q , XUAN Y B , et al. Maneuver decision of autonomous air combat of unmanned combat aerial vehicle based on deep neural network[J]. Acta Armamentarii, 2020, 41 (8): 1613- 1622. doi: 10.3969/j.issn.1000-1093.2020.08.016
18	ZHU J D , FU X W , QIAO Z . UAVs maneuver decision-making method based on transfer reinforcement learning[J]. Computational Intelligence and Neuroscience, 2022, 2022 (1): 2399796.
19	章胜, 周攀, 何扬, 等. 基于深度强化学习的空战机动决策试验[J]. 航空学报, 2023, 44 (10): 122- 135.
	ZHANG S , ZHOU P , HE Y , et al. Air combat maneuver decision-making test based on deep reinforcement learning[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44 (10): 122- 135.
20	杨晟琦, 田明俊, 司迎利, 等. 基于分层强化学习的无人机机动决策[J]. 火力与指挥控制, 2023, 48 (8): 48- 52.
	YANG S Q , TIAN M J , SI Y L ,et al. Research on UAV maneuver decision-making based on hierarchical reinforcement learning[J]. Fire Control & Command Control, 2023, 48 (8): 48- 52.
21	ERNEST N , COHEN K , KIVELEVITCH E , et al. Genetic fuzzy trees and their application towards autonomous training and control of a squadron of unmanned combat aerial vehicles[J]. Unmanned Systems, 2015, 3 (3): 185- 204.
22	BLUMHAGEN R Z , SCHWARTZ D A . Identification of influe- ntial rare variants in aggregate testing using random forest importance measures[J]. Annals of Human Genetics, 2023, 87 (4): 184- 195.
23	程绍驰, 游光荣. 基于改进随机森林算法的评估指标精简方法研究[J]. 系统工程与电子技术, 2023, 45 (7): 2108- 2113. doi: 10.12305/j.issn.1001-506X.2023.07.21
	CHENG S C , YOU G R . Research on the method of simplifying evaluation index based on improved random forest algorithm[J]. Systems Engineering and Electronics, 2023, 45 (7): 2108- 2113. doi: 10.12305/j.issn.1001-506X.2023.07.21
24	DANIEL J L , ZHU R Q , CUI Y F , et al. Dimension reduction forests: local variable importance using structured random fo-rests[J]. Journal of Computational and Graphical Statistics, 2022, 31 (4): 1104- 1113.
25	CHEN T Q, CARLOS G. XGBoost: a scalable tree boosting system[C]// Proc. of the 22nd ACM SIGKDD International Conference on Knowledge Giscovery and Data Mining, 2016.
26	VADHWANI D Y , DEVENDRA T . Prediction of extent of da-mage in vehicle during crash using improved XGBoost model[J]. International Journal of Crashworthiness, 2023, 28 (3): 299- 305.
27	REN Q X , WANG J G . Research on enterprise digital-level classification based on XGBoost model[J]. Sustainability, 2023, 15 (3): 2699.
28	王琳蒙. 非完备信息下的无人机空战动态博弈决策研究[D]. 南京: 南京航空航天大学, 2023.
	WANG L M. Research on dynamic game decision-making of UAV's air combat under incomplete information[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2023.
29	刘昊天. 无人机空战对抗博弈决策研究[D]. 南京: 南京航空航天大学, 2022.
	LIU H T. Research on game decision of UAV air combat confrontation[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2022.
30	STROBL C , BOULESTEIX A L , KNEIB T , et al. Conditional variable importance for random forests[J]. BMC Bioinformatics, 2008, 9 (1): 307.

输入特征	编号
(x_A, y_A, z_A)	e₁
V_A	e₂
θ_A	e₃
Ψ_A	e₄
(x_B, y_B, z_B)	e₅
V_B	e₆
θ_B	e₇
Ψ_B	e₈

输出属性	编号
匀速属性	w₁
左转弯	w₂
右转弯	w₃
加速	w₄
减速	w₅
俯冲	w₆
爬升	w₇

编号	x_A	y_A	z_A	V_A	θ_A	Ψ_A	x_B	y_B	z_B	V_B	θ_B	Ψ_B
4	53	51	2.3	197	30	4.1	25	7	6.15	150	10	13.3
4	35	11	7.73	170	4	15.2	3	35	8.58	158	0	16
3	17	7	5.12	198	3	9.6	50	51	6.99	178	20	1.3
6	42	12	11.55	178	16	13.6	43	10	4.98	178	19	18.5
2	2	59	6.75	155	27	12.5	55	23	8.96	177	12	11.4

编号	y_A	z_A	θ_A	x_B	y_B	z_B	V_B	Ψ_B
4	36	4.58	25	60	7	7.63	157	11.1
4	8	8.70	10	29	11	1.91	145	4.9
3	15	7.02	5	33	48	8.83	179	4.8
6	41	8.66	7	48	40	5.68	178	5.1
2	29	5.23	1	3	13	5.90	176	2.6

算法处理	模型平均训练时间/s
初始数据集训练模型	4.493
RF特征降维后训练模型	3.515