基于差分窗口生成式对抗网络的空战态势评估

doi:10.12305/j.issn.1001-506X.2024.08.21

摘要/Abstract

摘要：

针对飞机在空战中采集的飞行参数数据成分复杂、标签存在缺失等问题, 提出一种基于生成式对抗网络(generative adversarial network, GAN)的半监督空战态势评估模型。首先根据各要素权重提取空战数据的主要影响因子, 随后进行差分化和窗口化处理, 利用差分方法将态势信息相对化为一维特征向量, 窗口化信息生成反映两架载机态势信息的特征矩阵, 并送入网络进行半监督训练。仿真结果表明, 该模型在样本标签缺失的情况下具有良好的态势分析效果, 对于4种态势的识别准确率达90.91%。

关键词: 态势评估, 半监督学习, 差分窗口, 生成式对抗网络

Abstract:

Aiming at the complex composition and missing tags of the flight reference data collected by the aircraft during the air combat, a semi-supervised air combat situation assessment model is proposed based on the generative adversarial network (GAN). Firstly, the main influencing factors of the air combat data are extracted according to the weights of each element, then the differencing and windowing processing are carried out. The situation information is relativeized into a one-dimensional feature vector using the differential method. The windowing information generates a feature matrix reflecting the situation information of the two carrier aircrafts, which is sent to the network for semi-supervised training. Simulation results show that the model has a good situation analysis effect in the case of sample labels missing, and the recognition accuracy of the four situations is 90.91%.

Key words: situation assessment, semi-supervised learning, differential window, generative adversarial network (GAN)

中图分类号:

方伟, 张婷婷, 谭凯文, 汤淼. 基于差分窗口生成式对抗网络的空战态势评估[J]. 系统工程与电子技术, 2024, 46(8): 2738-2746.

Wei FANG, Tingting ZHANG, Kaiwen TAN, Miao TANG. Air combat situation assessment based on differential window generative adversarial network[J]. Systems Engineering and Electronics, 2024, 46(8): 2738-2746.

图/表 13

表1

图1

图2

图3

表2

表3

图4

图5

图6

表4

图7

图8

表5

参考文献 30

1	谷玉荣, 黄耀雄, 高艳, 等. 基于K-Means聚类算法的空中态势威胁挖掘[J]. 火力与指挥控制, 2019, 44 (4): 92- 96. doi: 10.3969/j.issn.1002-0640.2019.04.019
	GU Y R , HUANG Y X , GAO Y , et al. Aerial posture threat mining based on K-Means clustering algorithm[J]. Fire Control and Command Control, 2019, 44 (4): 92- 96. doi: 10.3969/j.issn.1002-0640.2019.04.019
2	游航航, 余敏建, 韩其松, 等. 基于攻击区的空战指挥引导态势评估[J]. 兵器装备工程学报, 2019, 40 (12): 29- 34. doi: 10.11809/bqzbgcxb2019.12.007
	YOU H H , YU M J , HAN Q S , et al. Air combat command based on the attack zone guides the posture assessment[J]. Journal of Weapons and Equipment Engineering, 2019, 40 (12): 29- 34. doi: 10.11809/bqzbgcxb2019.12.007
3	胡涛, 王栋, 姜龙亭, 等. 空战智能态势评估技术研究与展望[J]. 军事文摘, 2020, (5): 19- 22.
	HU T , WANG D , JIANG L T , et al. Research and prospect of intelligent posture assessment technology for air comba[J]. Mili- tary Digest Journal, 2020, (5): 19- 22.
4	史振庆, 梁晓龙, 张佳强, 等. 基于导弹攻击区的空战态势评估[J]. 火力与指挥控制, 2018, 43 (9): 89-93, 98. doi: 10.3969/j.issn.1002-0640.2018.09.019
	SHI Z Q , LIANG X L , ZHANG J Q , et al. Assessment of the air combat posture based on the missile attack zone[J]. Fire Control and Command Control, 2018, 43 (9): 89-93, 98. doi: 10.3969/j.issn.1002-0640.2018.09.019
5	KUMAR S, TRIPATHI B K. Modelling of threat evaluation for dynamic targets using Bayesian network approach[C]//Proc. of the International Conference on Emerging Trends in Engineering, Science and Tecnology, 2016: 1268-1275.
6	肖亮, 黄俊, 徐钟书. 基于空域划分的超视距空战态势威胁评估[J]. 北京航空航天大学学报, 2013, 39 (10): 1309- 1313.
	XIAO L , HUANG J , XU Z S . An assessment of the threat of an over-the-horizon air combat posture based on the division of airspace[J]. Journal of Beijing University of Aeronautics and Astronautics, 2013, 39 (10): 1309- 1313.
7	程天发, 葛泉波, 陈哨东, 等. 基于改进空战威胁评估模型的权重计算方法比较[J]. 火力与指挥控制, 2016, 41 (1): 32- 36. doi: 10.3969/j.issn.1002-0640.2016.01.008
	CHENG T F , GE Q B , CHEN S D , et al. Comparison of weight calculation methods based on improved air combat threat assessment model[J]. Fire Control and Command Control, 2016, 41 (1): 32- 36. doi: 10.3969/j.issn.1002-0640.2016.01.008
8	PATIL K, NAGWANI N K, TRIPATHI S. A parametric study of partitioning and density based clustering techniques for boxplot generation[C]//Proc. of the 3rd International Conference for Convergence in Technology, 2018.
9	SUN G L, LANG F, YANG M M, et al. Application protocols identification using non-parametric estimation method[C]//Proc. of the 6th International Forum on Strategic Technology, 2011: 765-768.
10	XU X M , YANG R N , FU Y . Situation assessment for air combat based on novel semi-supervised naive Bayes[J]. Journal of Systems Engineering and Electronics, 2018, 29 (4): 768- 779. doi: 10.21629/JSEE.2018.04.11
11	JI H, HAN Q, LI X Y, et al. Air combat situation assessment based on improved cloud model theory[C]//Proc. of the 8th Joint International Information Technology and Artificial Intelligence Conference, 2019: 754-758.
12	LI F, LIU J B, ZHANG F H, et al. Situation assessment for beyond-visual-range air combat situation assessment based on dynamic Bayesian network[C]//Proc. of the 25th Chinese Control and Decision Conference, 2013: 591-594.
13	ZHAO K X, HUANG C Q. Air combat situation assessment for UAV based on improved decision tree[C]//Proc. of the Chinese Control and Decision Conference, 2018: 1772-1776.
14	XU X M, YANG R N, YANG Y. Threat assessment in air combat based on ELM neural network[C]//Proc. of the International Conference on Artificial Intelligence and Computer App- lications, 2019: 114-120.
15	LU C G, ZHOU Z L, LIU H Q, et al. Situation assessment of far-distance attack air combat based on mixed dynamic Bayesian networks[C]//Proc. of the 37th Chinese Control Conference, 2018: 4569-4574.
16	孙永芹, 世文荣, 彭海军, 等. 基于改进非参量法的现代空战威胁评估研究[J]. 火控雷达技术, 2013, 42 (3): 28- 33.
	SUN Y Q , SHI W Y , PENG H J , et al. Modern air combat threat assessment research based on improved non-parametric method[J]. Journal of Fire Control Radar Technology, 2013, 42 (3): 28- 33.
17	ARORA D, GARG M, GUPTA M. Diving deep in deep convolutional neural network[C]//Proc. of the 2nd International Conference on Advances in Computing, Communication Control and Networking, 2020: 749-751.
18	ÖZMEN E, õOGUN F, ALTIPARMAK F. Classification of imbalanced dataset using generative adversarial nets[C]//Proc. of the 28th Signal Processing and Communications Applications Conference, 2020.
19	TAHMID M, ALAM S, AKRAM M K. Comparative analysis of generative adversarial networks and their variants[C]//Proc. of the 23rd International Conference on Computer and Information Technology, 2020.
20	GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[C]//Proc. of the Neural Information Process System, 2014: 2672-2680.
21	LI X Y, ZHANG Z K. The comparison between conditional generative adversarial nets and deep convolutional generative adversarial network, and its GUI-related application[C]//Proc. of the 2nd International Conference on Big Data & Artificial Intelligence & Software Engineering, 2021: 601-609.
22	RADFORD A, METZ L, CHINTALA S. Unsupervised representation learning with deep convolutional generative adversarial networks[EB/OL]. [2022-06-20]. https://arxiv.org/pdf/1511.06434.pdf.
23	韩博, 周一鹏, 钱程, 等. 基于深度卷积生成对抗神经网络的空战态势评估[C]//2021年无人系统高峰论坛, 2021: 43-49.
	HAN B, ZHOU Y P, QIAN C, et al. Air combat posture assessment based on deep convolutional generation of adversarial neural networks[C]//Proc. of the 2021 Unmanned Systems Summit Forum, 2021: 43-49.
24	YAN P Y , HE F , YANG Y J , et al. Semi-supervised representation learning for remote sensing image classification based on generative adversarial networks[J]. IEEE Access, 2020, 8, 54135- 54144. doi: 10.1109/ACCESS.2020.2981358
25	王坤峰, 苟超, 段艳杰, 等. 生成式对抗网络GAN的研究进展与展望[J]. 自动化学报, 2017, 43 (3): 321- 332.
	WANG K F , GOU C , DUAN Y J , et al. Research progress and prospect of generative adversarial network GAN[J]. Journal of Automation, 2017, 43 (3): 321- 332.
26	林懿伦, 戴星原, 李力, 等. 人工智能研究的新前线: 生成式对抗网络[J]. 自动化学报, 2018, 44 (5): 775- 792.
	LIN Y L , DAI X Y , LI L , et al. A new frontier in AI research: generative adversarial networks[J]. Journal of Automation, 2018, 44 (5): 775- 792.
27	王万良, 李卓蓉. 生成式对抗网络研究进展[J]. 通信学报, 2018, 39 (2): 135- 148.
	WANG W L , LI Z R . Research progress on generative adversarial networks[J]. Journal of Communications, 2018, 39 (2): 135- 148.
28	TAN K W , YAN W J , ZHANG L M , et al. Semi-supervised specific emitter identification based on bispectrum feature extraction CGAN in multiple communication scenarios[J]. IEEE Trans.on Aerospace and Electronic Systems, 2023, 59 (1): 292- 310. doi: 10.1109/TAES.2022.3184619
29	SUN S L, HUANG R Q. An adaptive K-nearest neighbor algorithm[C]//Proc. of the 7th International Conference on Fuzzy Systems and Knowledge Discovery, 2010: 91-94.
30	LU C G, ZHOU Z L, LIU H Q, et al. Situation assessment of far-distance attack air combat based on mixed dynamic Bayesian networks[C]//Proc. of the 37th Chinese Control Conference, 2018: 4569-4574.

参数	取值范围
两机相对距离ΔR/km	[0, 20]
两机相对速度ΔV/(km/h)	[0, 500]
红方航向角φ₁/rad	[0, π]
蓝方航向角φ₂/rad	[0, π]
红方俯仰角δ₁/rad	[0, π]
蓝方俯仰角δ₂/rad	[0, π]
红方横滚角γ₁/rad	[0, π]
蓝方横滚角γ₂/rad	[0, π]

层	输出尺寸	层	输出尺寸
input_1	(1, 80)	batch _2	(1, 4, 128)
reshape	(1, 8, 10)	Activation_2	(1, 4, 128)
Conv2d	(1, 4, 64)	Conv2d_4	(1, 4, 128)
Conv2d_1	(1, 4, 128)	batch _3	(1, 4, 128)
batch	(1, 4, 128)	Activation_3	(1, 4, 128)
activation	(1, 4, 128)	Flatten	(512)
Conv2d_2	(1, 4, 128)	dense	(1 024)
batch _1	(1, 4, 128)	dense_1	(512)
Activation_1	(1, 4, 128)	dense_2	(64)
Conv2d_3	(1, 4, 128)	Softmax/Sigmoid	4/1

层	输出尺寸	层	输出尺寸
input_3	(1, 100)	Leaky _2	(2, 256, 128)
Dense_8	(1, 8 192)	batch _10	(2, 256, 128)
Reshape_2	(1, 128, 64)	Conv2d_13	(2, 256, 64)
Conv2d_10	(1, 128, 64)	Leaky_3	(2, 256, 64)
Leaky_re_lu	(1, 128, 64)	batch _11	(2, 256, 64)
batch _8	(1, 128, 64)	Flatten_3	(32 768)
Up_sam2d	(2, 256, 64)	dense_9	(1 024)
Conv2d_11	(2, 256, 128)	dense_10	(512)
Leaky _1	(2, 256, 128)	dense_11	(256)
batch _9	(2, 256, 128)	dense_12	(80)
Conv2d_12	(2, 256, 128)	Reshape_3	(1, 80)

训练样本占比	实验次数
训练样本占比	1	2	3	4	平均值
0.5	80.84	78.68	80.52	82.47	80.63
0.6	83.77	80.62	81.17	84.42	82.50
0.7	83.31	83.99	80.52	82.79	82.65
0.8	82.14	83.44	87.01	86.04	84.66
0.9	80.52	86.69	87.34	88.96	85.88

算法	识别率
传统神经网络模型	85.63
自适应KNN算法模型	83.66
动态贝叶斯网络模型	85.47
基于差分窗口的GAN模型	90.91