数据和知识双驱动的空中集群目标作战意图识别

doi:10.12382/bgxb.2024.0113

摘要/Abstract

摘要：

针对集群目标空间特性多元时变和传统数据驱动模型过分依赖经验样本等问题,提出一种针对集群目标的数据和知识双驱动作战意图识别方法。考虑集群目标空间形态等编队特点,构造基于目标编队外包络线和最小外接矩形的集群特征向量,增强敌情数据的特征表达效果;建立基于专家经验的知识模型和结合注意力机制的长短期记忆(Long short-term memory,LSTM)网络模型,基于专家经验的知识模型根据约束规则生成意图预识别向量,LSTM模型预测输出意图概率分布的残差;利用一种可学习的残差估计器结构,自适应调整双模型的融合比率,并设计多目标损失函数控制双模型的影响权重,最终通过双模型的融合有效克服传统数据模型高精度和数据样本不足的矛盾。实验表明,提出方法的精度相比LSTM和Attention-LSTM分别提升约5.34%和4.98%,且对样本量的依赖性显著低于传统数据驱动方法。

关键词: 集群目标, 作战意图, 数据驱动, 知识驱动, 注意力机制

Abstract:

Aiming at the diverse spatiotemporal characteristics of cluster targets and the excessive reliance of traditional data driven models on empirical samples,this paper proposes an algorithm for combat intent recognition driven by both data and knowledge.A cluster feature vector based on the virtual envelope and minimum bounding rectangle of target formation are constructed to enhance the feature expression of enemy situation data,which takes the cluster characteristics,such as the spatial form of cluster targets,into account.A knowledge model based on military expert experience and a long short-term memory (LSTM) network model with attention mechanism are established then.The knowledge model generates the intent pre-recognition vectors based on constraint rule,while the LSTM network model predicts the residual of intent probability distribution.The fusion ratio of both models is adaptively adjusted by utilizing a learnable residual estimator structure.A multi-objective loss function is designed to control the influence weights of the dual models.Ultimately,the fusion of the dual models overcomes the contradiction between the high accuracy of traditional data models and the insufficient data samples.Experimental results indicate that the proposed method improves the recognition accuracy to about 5.34% and 4.98% compared to LSTM and Attention-LSTM,respectively,and has significantly lower dependence on sample size than traditional data-driven methods.

Key words: cluster targets, combat intent, data driving, knowledge driving, attention mechanism

李洋军, 黄琦龙, 杨力, 陈旭. 数据和知识双驱动的空中集群目标作战意图识别[J]. 兵工学报, 2025, 46(2): 240113-.

LI Yangjun, HUANG Qilong, YANG Li, CHEN Xu. Combat Intention Recognition of Air Cluster Targets Driven by Data and Knowledge[J]. Acta Armamentarii, 2025, 46(2): 240113-.

图/表 14

表1 考虑的目标特征

Table 1 Target features considered in this paper

运动学特征	状态特征	电磁频谱特征^[12]
水平位置	方位角	脉冲宽度
速度	航向角	脉冲重复频率
加速度	高度	载波频率

表2 典型无人战斗机的突防意图模板

Table 2 Template for penetration intention of typical unmanned fighter

参数	速度/(m·s^-1)	加速度/ (m·s^-2)	高度/m	航向变化率/(°)	脉冲宽度/μs	脉冲重复频率/MHz	载波频率/GHz	意图
典型值	30~50	0.15~1.5	300~1600	0~15	4~8	5~9	1.5~4.5	突防
允许值	20~90	0~1.6	250~2000	0~50	0~12	0~11	0~7.5	突防

图1 基于模板匹配的知识驱动意图识别

Fig.1 Knowledge driven intention recognition based on template matching

图2 数据和知识双驱动作战意图识别框架

Fig.2 Data and knowledge-driven combat intent recognition framework

图3 双驱动模型的训练和识别流程

Fig.3 Training and recognition process of data and knowledge-driven model

表3 模型消融实验

Table 3 Model ablation experiment

LSTM	注意力机制	集群特征	残差估计器	准确率/%	标准差/%
P				91.76	0.62
P	P			92.12	0.47
P		P		95.68	0.28
P	P	P		96.32	0.25
P	P	P	P	97.10	0.19

图4 不同模型的性能比较

Fig.4 Performance comparison among several models

图5 不同样本量下的模型识别精度

Fig.5 Model recognition accuracy under different sample sizes

图6 软性损失系数β、样本量N和精度的关系

Fig.6 Relationship among soft loss coefficient β,sample size N and accuracy

图7 作战想定下的目标运动轨迹

Fig.7 Target movement trajectory under combat scenario

表4 作战想定描述

Table 4 Description of combat scenarios

阶段	作战情形	实际意图
(a)~(b)	蓝方目标保持纵队并开始接近红方基地A。	-
(b)~(c)	该目标在红方基地A的上空徘徊飞行并实施电子干扰,以配合蓝方的其余空中作战力量对基地A实施打击。	电子干扰
(c)~(d)	红方基地A在(c)时刻被摧毁,红方开始在基地B上空部署一些空中防御兵力。该蓝方目标变换队形为菱形,欲缓缓下降高度并靠近红方基地B,发动佯攻战术以吸引红方的攻击火力。	佯攻
(d)之后	该目标在(d)时刻开始加速飞行,试图配合蓝方的其余空中作战力量共同对红方基地B实施打击,飞行一段时间后开启雷达追踪模式,转弯朝向红方空中兵力发起突袭进攻。	攻击

图8 DM的识别结果

Fig.8 The recognition results of DM

图9 KM的识别结果

Fig.9 The recognition results of KM

图10 提出模型的识别结果

Fig.10 The recognition results of the proposed model

参考文献 22

[1]	李伟生, 王宝树. 实现态势估计的一种模板匹配算法[J]. 计算机科学, 2006, 33(5):229-230,249.
	LI W S, WANG B S. A template matching algorithm for implementing situation assessment[J]. Computer Science, 2006, 33(5):229-230,249. (in Chinese)
[2]	吉祥. 基于聚类分群与模板匹配的态势估计方法研究[D]. 国防科学技术大学, 2016.
	JI X. Study of situation assessment techniques based on clustering and matching[D]. National University of Defense Technology, 2016. (in Chinese)
[3]	李香亭. 态势估计中目标意图识别的研究与实现[D]. 中北大学, 2012.
	LI X T. The research and implementation of situation assessment in the target intention recognition[D]. North University of China, 2012. (in Chinese)
[4]	赵福均, 周志杰, 胡昌华. 基于置信规则库和证据推理的空中目标意图识别方法[J]. 电光与控制, 2017, 24(8):15-19.
	ZHAO F J, ZHOU Z J, HU C H, et al. Aerial target intention recognition approach based on belief-rule-base and evidential reasoning[J]. Electronics Optics & Control, 2017, 24(8):15-19. (in Chinese)
[5]	蒲志强, 易建强, 刘振. 知识和数据协同驱动的群体智能决策方法研究综述[J]. 自动化学报, 2022, 48(3):627-643.
	PU Z Q, YI J Q, LIU Z. Knowledge-based and data-driven integrating methodologies for collective intelligence decision making:a survey[J]. Journal of Automatica Sinica, 2022, 48(3):627-643. (in Chinese)
[6]	乔殿峰, 梁彦, 马超雄. 多域作战下的群目标意图识别与预测[J]. 系统工程与电子技术, 2022, 44(11):3403-3412. doi: 10.12305/j.issn.1001-506X.2022.11.15
	QIAO D F, LIANG Y, MA C X, et al. Recognition and prediction of group target intention in multi-domain operations[J]. Systems Engineering and Electronics, 2022, 44(11):3403-3412. (in Chinese) doi: 10.12305/j.issn.1001-506X.2022.11.15
[7]	JING Q, GUO X T, JIN W D, et al. Intention recognition of aerial targets based on Bayesian optimization algorithm[C]// Proceedings of 2017 2nd IEEE International Conference on Intelligent Transportation Engineering. Singapore:IEEE, 2017:356-359.
[8]	YANG Z, SUN Z X, PIAO H Y, et al. Online hierarchical recognition method for target tactical intention in beyond-visual-range air combat[J]. Defense Technology, 2022, 18(8):1349-1361.
[9]	李战武, 李双庆, 彭明毓. 基于注意力机制改进的LSTM空战目标意图识别方法[J]. 电光与控制, 2023, 30(3):1-7.
	LI Z W, LI S Q, PENG M Y. Air combat intention recognition method of target based on LSTM improved by attention mechanism[J]. Electronics Optics & Control. 2023, 30(3):1-7. (in Chinese)
[10]	李峰, 王琦, 胡健雄. 数据与知识联合驱动方法研究进展及其在电力系统中应用展望[J]. 中国电机工程学报, 2021, 41(13): 4377-4390.
	LI F, WANG Q, HU J X, et al. Combined data-driven and knowledge-driven methodology research advances and its applied prospect in power systems[J]. Proceedings of the CSEE. 2021, 41(13): 4377-4390. (in Chinese)
[11]	葛顺. 基于规则发现和贝叶斯推理的战术意图识别技术[D]. 哈尔滨工程大学, 2015.
	GE S. Research on tactical intention recognition based on rule discovery and Bayesian reasoning[D]. Harbin Engineering University, 2015. (in Chinese)
[12]	张建廷, 周万宁, 晏谢飞. 基于LSTM和模糊推理的海上目标行为意图智能预测方法[J]. 中国电子科学研究院学报, 2022, 17(9):897-904.
	ZHANG J T, ZHOU W N, YAN X F, et al. Intelligent prediction for behavioral intent of marine targets based on LSTM and Fuzzy Reasoning[J]. Journal of China Academy of Electronics and Information Technology, 2022, 17(9):897-904. (in Chinese)
[13]	余源丰. 群目标态势推理关键技术研究[D]. 西安电子科技大学, 2020.
	YU Y F. Research on key technologies of group targets situation reasoning[D]. Xidian University, 2020. (in Chinese)
[14]	陈刚, 姚丽亚, 王国新. 面向鲁棒决策的战场态势评估人机共识形成方法[J]. 兵工学报, 2022, 43(11):2953-2964. doi: 10.12382/bgxb.2021.0557
	CHEN G, YAO L Y, WANG G X, et al. A human-machine consensus formation method for robust decision making in battlefield situation assessment[J]. Acta Armamentarii, 2022, 43(11): 2953-2964. (in Chinese) doi: 10.12382/bgxb.2021.0557
[15]	ZHANG Y, HUANG F H, DENG X Y, et al. Air target intention recognition and causal effect analysis combining uncertainty information reasoning and potential outcome framework[J]. Chinese Journal of Aeronautics, 2024, 37(1):287-299.
[16]	BAHNSEN A C, AOUADA D, STOJANOVIC A, et al. Feature engineering strategies for credit card fraud detection[J]. Expert Systems with Applications, 2016, 51:134-142.
[17]	PEAD E, GIARRATANO Y, TATHAM A J, et al. 2D alpha-shapes to quantify retinal microvasculature morphology and their application to proliferative diabetic retinopathy characterisation in fundus photographs[J]. Scientific Reports, 2021, 11(1):22814. doi: 10.1038/s41598-021-02329-5 pmid: 34819594
[18]	彭正初. 基于傅里叶描述子的物体形状识别的研究[D]. 哈尔滨工业大学, 2017.
	PENG Z C. Research on object shape recognition based on Fourier descriptors[D]. Harbin Institute of Technology, 2017. (in Chinese)
[19]	郑志伟, 管雪元, 傅健. 基于卷积神经网络与长短期记忆神经网络的弹丸轨迹预测[J]. 兵工学报, 2023, 44(10):2975-2983.
	ZHENG Z W, GUAN X Y, FU J, et al. Projectile trajectory prediction based on CNN-LSTM model[J]. Acta Armamentarii, 2023, 44(10):2975-2983. (in Chinese) doi: 10.12382/bgxb.2022.0511
[20]	BAHARI M, NEJJAR I, ALAHI A. Injecting knowledge in data-driven vehicle trajectory predictors[J]. Transportation Research Part C:Emerging Technologies, 2021, 128:103010.
[21]	周志华. 机器学习[M]. 北京: 清华大学出版社, 2016.
	ZHOU Z H. Machine learning[M], Beijing: Tsinghua University Press, 2016. (in Chinese)
[22]	DAW A, KARPATNE A, WATKINS W D, et al. Physics-guided neural networks (PGNN):An application in lake temperature modeling[M]//Knowledge Guided Machine Learning. Chapman and Hall/CRC, 2022:353-372.