Research on Projectile Penetration Target Material Detection Using Self-attention and Graph Neural Networks

FU Shijie; SHAO Weiping; HAO Yongping

doi:10.15892/j.cnki.djzdxb.2025.03.017

Journal of Projectiles, Rockets, Missiles and Guidance >

2025 , Vol. 45 >Issue 3: 386 - 391

DOI: https://doi.org/10.15892/j.cnki.djzdxb.2025.03.017

Research on Projectile Penetration Target Material Detection Using Self-attention and Graph Neural Networks

FU Shijie ^,¹^,² ,
SHAO Weiping ^,¹^,² ,
HAO Yongping ¹^,²

Expand

¹ School of Mechanical Engineering,Shenyang Ligong University, Shenyang 110159, Liaoning, China
² Liaoning Key Laboratory of Advanced Manufacturing Technology and Equipment, Shenyang 110159, Liaoning, China

Received date: 2024-08-28

Online published: 2025-07-09

Fold

Abstract

In the context of modern intelligent warfare, achieving precision strikes with artillery and accurately assessing shooting targets presents significant challenges. To address these challenges, a novel detection model that integrates graph neural networks (GNNs) with a self-attention mechanism is proposed for analyzing multivariate time series data. The model utilizes real-time data from multiple sensors embedded within the projectile to classify and detect the material of the target upon impact. This enables a precise determination of whether the projectile has struck the intended target, complementing other detection methods.By incorporating the graph model, the self-attention mechanism facilitates message passing across sensor dimensions, thereby improving the model's ability to detect variations in data from different sensors and prioritize relevant and effective information. To address the complexities of irregular data sampling and the intricate nature of multivariate time series, the model leverages temporal information relationships to construct a temporal self-attention mechanism network. This network is specifically designed to model complex time series data by capturing temporal dependencies.Experimental results indicate that the proposed algorithm outperforms mainstream models in terms of accuracy, with a material recognition rate for target plates as high as 90%, significantly improving precision in target impact assessment. This study provides a crucial reference for the development of intelligent detection systems for target strike identification, contributing to the advancement of more accurate and reliable targeting methods in modern warfare.

Key words： graph neural network; time series; self-attention mechanism; LS-DYNA; classification

Cite this article

FU Shijie , SHAO Weiping , HAO Yongping . Research on Projectile Penetration Target Material Detection Using Self-attention and Graph Neural Networks[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2025 , 45(3) : 386 -391 . DOI: 10.15892/j.cnki.djzdxb.2025.03.017

[an error occurred while processing this directive]

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	连细南, 陈栋, 石胜斌. 炮弹炸点坐标检测技术研究综述[J]. 兵器装备工程学报, 2022, 43(1): 35-41. LIAN X, CHEN D, SHI S B. Survey on coordinate detection technology for burst point of artillery shells[J]. Journal of Ordnance Equipment Engineering, 2022, 43(1): 35-41.

[2]	HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]// IEEE.IEEE Conference on Computer Vision and Pattern Recognition, June 27-30, 2016, Piscataway, State of New Jersey, New York: IEEE, 2016: 770-778.

[3]	HE K, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]// International Conference on Computer Vision. October 22-27, 2017, Piscataway, NJ, New York: IEEE, 2017: 2980-2988.

[4]	NAM H, BAEK M, HAN B. Modeling and propagating cnns in a tree structure for visual tracking[EB/OL]. [2024-08-28].

[5]	BRODLEY C E, UTGOFF P E. Multivariate decision trees[J]. Machine Learning, 1995, 19(1): 45-77.

[6]	SHAWE-TAYLOR J, SUN S. A review of optimization methodologies in support vector machines[J]. Neurocomputing, 2011, 74(17): 3609-3618.

[7]	DE AMORIM R C. A survey on feature weighting based K Means algorithms[J]. Jounal of Classification, 2016, 33: 210-242.

[8]	CUN Y L, BOSER B, DENKER J S, et al. Handwritten digit recognition with a back-propagation network[J]. Advances in Neural Information Processing Systems, 1997, 2(2): 396-404.

[9]	任利强, 贾舒宜, 王海鹏, 等. 基于深度学习的时间序列分类研究综述[J]. 电子与信息学报, 2024, 46(8): 3094-3116. REN L Q, JIA S Y, WANG H P, et al. A review of research on time series classification based on deep learning[J]. Journal of Electronics & Information Technology, 2024, 46(8): 3094-3116.

[10]	LEE S M, YOON S M, CHO H. Human activity recognition from accelerometer data using convolutional neural network[C]//2017 IEEE International Conference on Big Data and Smart Computing (BigComp), February 13-16, 2017, Jeju Island, South Korea, New York: IEEE, 2017: 131-134.

[11]	SINGH S P, SHARMA M K, LAY-EKUAKILLE A, et al. Deep ConvLSTM with self-attention for human activity decoding using wearable sensors[J]. IEEE Sensors Journal, 2021, 21(6): 8575-8582.

[12]	JIANG W, LUO J. Graph neural network for traffic forecastting: A survey[J]. Expert Systems with Applications, 2022, 207: 117921.

[13]	ZHANG X, ZEMAN M, TSILIGKARIDIS T, et al. Graph-guided network for irregularly sampled multivariate time series[EB/OL]. [2024-08-28].

[14]	DENG A, HOOI B. Graph neural network-based anomaly detection in multivariate time series[C]// Proceedings of the AAAI Conference on Artificial Intelligence, February 2-9, 2021, New York, California: AAAI, 2021, 35(5): 4027-4035.

[15]	CHENG Z, YANG Y, WANG W, et al, Time2graph: Revisiting time series modeling with dynamic shapelets[C]// Proceedings of the AAAI Conference on Artificial Intelligence, February 7-12, 2020, New York, California: AAAI, 2020, 34(4): 3617-3624.

[16]	CHENG Z, YANG Y, JIANG S, et al. Time2graph+: Bridging time series and graph repressentation learning via multiple attentions[J]. IEEE Transactions on Knowledge and Data Engineering, 2021, 35(2): 2078-2090.

[17]	ZHA D, LAI K H, ZHOU K, et al. Towards similarity-awaretime-series classification[C]// Proceedings of the 2022 SIAM International Conference on Data Mining. Society of Industrial and Applied Mathematics, April 28-30, 2022, Alexandria, Virginia, USA, Philadelphia: SIAM, 2022: 199-207.

[18]	魏欣, 陈博, 张德志, 等. 刚性动能弹侵彻混凝土过载特性的数值模拟[J]. 现代应用物理, 2021, 12(4): 137-143. WEI X, CHEN B, ZHANG D Z, et al. Numerical simulation of overload characteristics of rigid kinetic energy projectile penetrating concret[J]. Modern Applied Physics. 2021, 12(4): 137-143.

[19]	张建海. 攻坚弹药子弹弹道环境的初步试验研究与数值分析[D]. 南京: 南京理工大学, 2006. ZHANG J H. Preliminary study experiment and numerical computation on submunition ballistic trajectory condition of hard structure munition[D]. Nanjing: Nanjing University of Science and Technology, 2006.

[20]	唐昌州, 智小琦, 高峰, 等. 钨球侵彻防弹衣加红松木复合靶的研究[J]. 爆炸与冲击, 2021, 41(6): 84-97. TANG C Z, ZHI X Q, GAO F, et al. Investigation on tungsten spheres penetrating into pine target covered with body armor[J]. Explosion and Shock Waves, 2021, 41(6): 84-97.

[21]	SMITH L N. Cyclical learning rates for training neural networks[C]//2017 IEEE winter conference on applications of computer vision (WACV), March 24-31, 2017, Santa Rosa, CA, New York: IEEE, 2017: 464-472.

Options

Outlines