基于人工神经网络的三棱柱定向装药结构破片初速预测模型

doi:10.12382/bgxb.2024.0346

摘要/Abstract

摘要：

棱柱形装药结构作为一种典型的非对称结构,其能量输出具有明显的方向性,预测其破片速度分布对于新型战斗部的结构设计和毁伤效率评估具有重要意义。针对棱柱形装药结构,建立一种基于人工神经网络的破片速度预测模型。为提高网络模型的效率和准确性,通过理论分析确定了影响破片速度分布的影响因素,为网络模型筛选出4个输入特征参数。通过调整这些特征参数的值,建立多组不同的数值模拟工况,通过数值模拟方法为网络模型提供数据集。用训练好的网络模型对测试集进行了预测,预测结果与数值模拟结果吻合较好,表明该网络模型预测棱柱形装药结构的破片分布具有较高的准确性,并且该神经网络模型具有良好的泛化能力。该神经网络模型具有计算速度快、预测精度高、易于建模等特点,可以较为精确地预测一端起爆条件下棱柱形结构的破片速度分布,为战斗部结构设计和毁伤效率评估提供数据参考。

关键词: 棱柱形壳体, 破片初速, 量纲分析, 人工神经网络

Abstract:

The prismatic charge structure,as a typical asymmetric structure,exhibits a pronounced directional energy output.It is important to predict the velocity distribution of its fragments for the structural design and damage efficiency assessment of warhead.For the prismatic charge structure,a fragment velocity prediction model based on an artificial neural network (ANN) is proposed.To improve the predictive efficiency and accuracy of the network model,the key factors affecting the fragment velocity distribution are identified through theoretical analysis,and four input characteristic parameters are selected for the network model.The multiple sets of different numerical simulation conditions are established by adjusting the values of these characteristic parameters,and a dataset is provided for the network model through numerical simulation method.The trained network model is used to predict the test set,and the predicted results are in good agreement with the numerically simulated results.The results indicate that the proposed network model has high accuracy in predicting the fragment distribution of prismatic charge structure,and has a good generalization capability.The neural network model is characterized by fast computation speed,high predictive accuracy,and easy-to-modeling.It can accurately predict the fragment velocity distribution of prismatic structure under single-end initiation conditions,thus providing important data support for the structural design and damage efficiency assessment of warheads.

Key words: prismatic casing, initial velocity, dimensional analysis, artificial neural network

宁建国, 汪齐, 栗建桥. 基于人工神经网络的三棱柱定向装药结构破片初速预测模型[J]. 兵工学报, 2025, 46(3): 240346-.

NING Jianguo, WANG Qi, LI Jianqiao. ANN-based Prediction Model for the Initial Velocity of Fragments in a Triangular Prism Directional Charge Structure[J]. Acta Armamentarii, 2025, 46(3): 240346-.

图/表 24

参考文献 27

[1]	WANG M F, LU F Y, LI X Y, et al. A formula for calculating the velocities of fragments from velocity enhanced warhead[J]. Propellants,Explosives,Pyrotechnics, 2013, 38(2):232-237.
[2]	DING L L, LI Z D, LIANG M Z, et al. The dispersion rule of fragments about the asymmetric shell[J]. Shock and Vibration, 2017, 2017:1-12.
[3]	MA T B, SHI X W, LI J, et al. Fragment spatial distribution of prismatic casing under internal explosive loading[J]. Defence Technology, 2020, 16 (4):910-921. doi: 10.1016/j.dt.2019.11.006
[4]	赵宇哲, 李健, 马天宝. 展开式定向战斗部展开过程实验研究[J]. 高压物理学报, 2016, 30(2):116-122.
	ZHAO Y Z, LI J, MA T B. Experiment on spread processes of the spreadable aimed warhead[J]. Chinese Journal of High Pressure Physics, 2016, 30(2):116-122. (in Chinese)
[5]	ZHAO Y Z, MA T B, LI J, et al. Study on dynamic response of multi-body structure under explosive driving[J]. Science China Technological Sciences, 2016, 59(9):1360-1369.
[6]	GURNEY R W. The initial velocities of fragments from bombs,shell and grenades[R]. Aberdeen Proving Ground,MD, US: Army Ballistic Research Lab., 1943.
[7]	ZULKOSKI T. Development of optimum theoretical warhead design criteria:B015617[R]. China Lake,CA, US: Naval Weapons Center, 1976:39-44.
[8]	LI W, HUANG G Y, FENG S S. Effect of eccentric edge initiation on the fragment velocity distribution of a cylindrical casing filled with charge[J]. International Journal of Impact Engineering, 2015, 80:107-115.
[9]	高月光, 冯顺山, 刘云辉, 等. 不同端盖厚度的圆柱形装药壳体破片初速分布[J]. 兵工学报, 2022, 43(7):1527-1536.
	GAO Y G, FENG S S, LIU Y H, et al. Initial velocity distribution of fragments from cylindrical charge shells with different thick end caps[J]. Acta Armamentarii, 2022, 43(7):1527-1536. (in Chinese) doi: 10.12382/bgxb.2021.0443
[10]	GUO Z W, HUANG G Y, LIU H, et al. Fragment velocity distribution of the bottom part of D-shaped casings under eccentric initiation[J]. International Journal of Impact Engineering, 2020, 144:103649.
[11]	NING J G, DUAN Y, XU X Z, et al. Velocity characteristics of fragments from prismatic casing under internal explosive loading[J]. International Journal of Impact Engineering, 2017, 109:29-38.
[12]	DENG Y X, ZHANG X F, XIAO C, et al. Experimental study on fragments dispersion characteristics of elliptical cross section casing under explosive loading[J]. Thin-Walled Structures, 2024, 197:111601.
[13]	SHI Y P, ZHOU T, GUO Z W, et al. Velocity distribution of preformed fragments from concave quadrangular charge structures[J]. International Journal of Impact Engineering, 2023, 176:104551.
[14]	LIU H, HUANG G Y, GUO Z W, et al. Fragments velocity distribution and estimating method of thin-walled cylindrical improvised explosive devices with different length-to-diameter ratios[J]. Thin-Walled Structures, 2022, 175:109212.
[15]	LI Y, YU C, SUO T, et al. Maximum fragment velocity of hollow charges with different aspect ratios[J]. International Journal of Impact Engineering, 2023, 178:104622.
[16]	YE P, DONG Y X, SUN Q T, et al. Double casing warhead with sandwiched charge:the axial distribution of fragments velocities[J]. Defence Technology, 2024, 34(4):201-216
[17]	LI Y, REN T F, WEN Y Q, et al. Impact of aspect ratio on fragment velocity distribution for hollow charges[J]. Materials & Design, 2023, 228:111815.
[18]	MARZOUK M, ELHAKEEM A, ADEL K. Artificial neural networks applications in construction and building engineering(1991-2021):science mapping and visualization[J]. Applied Soft Computing, 2024, 152:111174.
[19]	AHMED S A, AKMAR M A, B HUSSIN H, et al. Analysis of machine learning models and data sources to forecast burst pressure of petroleum corroded pipelines:a comprehensive review[J]. Engineering Failure Analysis, 2024, 155:107747.
[20]	肖师云, 冯成良, 陈文, 等. 基于BP神经网络的破片聚焦曲线优化[J]. 兵工自动化, 2022, 41(1):43-47.
	XIAO S Y, FENG C L, CHEN W, et al. Optimization of fragment focusing curve based on bp neural network[J]. Ordnance Industry Automation, 2022, 41(1):43-47. (in Chinese)
[21]	辛大钧, 薛琨. 基于人工神经网络的非球形破片阻力系数预测模型[J]. 兵工学报, 2022, 43(5):1083-1092. doi: 10.12382/bgxb.2021.0121
	XIN D J, XUE K. Artificial neural network-based prediction model for the air drag coefficient of non-spherical fragments[J]. Acta Armamentarii, 2022, 43(5):1083-1092. (in Chinese) doi: 10.12382/bgxb.2021.0121
[22]	杨芮, 王亚辉, 刘宇峰, 等. 预控破片战斗部周向刻槽参数正交优化分析[J]. 弹箭与制导学报, 2020, 40(6):54-57,61.
	YANG R, WANG Y H, LIU Y F, et al. Orthogonal optimization analysis of the parameters of the circumferential groove of the pre-control fragment warhead[J]. Journal of Projectiles,Rockets,Missiles and Guidance, 2020, 40(6):54-57,61. (in Chinese)
[23]	彭正午, 张庆, 王晓鸣, 等. 刻槽参数对预控破片形成情况的影响研究[J]. 火工品, 2013(1):17-20.
	PENG Z W, ZHANG Q, WANG X M, et al. The effectof groove parameters on the formation of preformed fragments[J]. Initiators & Pyrotechnics, 2013(1):17-20. (in Chinese)
[24]	李兴隆, 陈科全, 路中华, 等. 装填系数对破甲杀伤复合战斗部威力影响的数值模拟[J]. 含能材料, 2019, 27(6):535-540.
	LI X L, CHEN K Q, LU Z H, et al. Numerical simulation of the influence of charge-shell mass ratio on the damage power of anti-armor and anti-personnel composite warhead[J]. Chinese Journal of Energetic Materials, 2019, 27(6):535-540. (in Chinese)
[25]	LIAO W, JIANG J W, MEN J B, et al. Effect of the end cap on the fragment velocity distribution of a cylindrical cased charge[J]. Defence Technology, 2021, 17(3):1052-1061. doi: 10.1016/j.dt.2020.06.024
[26]	KETKAR N. Deep learning with python[M]. New York,NY, US: Manning Publications Co.,Shelter Island, 2017.
[27]	段妍. 爆炸加载刻槽平板结构动态响应研究[D]. 北京: 北京理工大学, 2018.
	DUAN Y. Dynamic response of grooved plate subjected to explosion loading[D]. Beijing: Beijing Institute of Technology, 2018. (in Chinese)

影响因素	物理量	量纲
炸药	m_e/g	M
	ρ_e/(g·cm^-3)	L^-3M
	E_e/(cm²·μs^-2)	L² T^-2
	γ_e/SI	1
铝壳	ρ_s/(g·cm^-3)	L^-3M
	E_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	μ_s/SI	1
	Y_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	h_s/cm	L
	S_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	R/cm	L
	n/cm	L
	α/SI	1
破片板	ρ_f/(g·cm^-3)	L^-3M
	E_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	μ_f/SI	1
	h_f/cm	L
	Y_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	S_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	H/cm	L
	W/cm	L
破片	x/cm	L
	y/cm	L
	l/cm	L
	破片速度v_f/(cm·μs^-1)	L T^-1

影响因素	物理量	量纲
炸药	m_e/g	M
	ρ_e/(g·cm^-3)	L^-3M
	E_e/(cm²·μs^-2)	L² T^-2
	γ_e/SI	1
铝壳	ρ_s/(g·cm^-3)	L^-3M
	E_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	μ_s/SI	1
	Y_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	h_s/cm	L
	S_s/(g·cm^-1·μs^-2)	L^-1M T^-2
	R/cm	L
	n/cm	L
	α/SI	1
破片板	ρ_f/(g·cm^-3)	L^-3M
	E_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	μ_f/SI	1
	h_f/cm	L
	Y_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	S_f/(g·cm^-1·μs^-2)	L^-1M T^-2
	H/cm	L
	W/cm	L
破片	x/cm	L
	y/cm	L
	l/cm	L
	破片速度v_f/(cm·μs^-1)	L T^-1

材料	A/MPa	B/MPa	C	n	m	D₁	D₂	D₃	D₄	D₅
45号钢	507	320	0.064	0.28	1.06	0.06	3.31	-1.96	0.001 8	0.58
LY-12	546	678	0.024	0.71	1.56	-0.068	0.451	-0.952	0.036	0.697

材料	A/MPa	B/MPa	C	n	m	D₁	D₂	D₃	D₄	D₅
45号钢	507	320	0.064	0.28	1.06	0.06	3.31	-1.96	0.001 8	0.58
LY-12	546	678	0.024	0.71	1.56	-0.068	0.451	-0.952	0.036	0.697

破片编号	数值模拟初速/ (m·s^-1)	试验初速/ (m·s^-1)	相对误差/%
1	1057.1	1144.3	-7.6
2	1155.6	1164.3	-0.8
3	1197.2	1182.0	1.3
4	1227.8	1196.9	2.6
5	1243.3	1208.8	2.9
6	1253.8	1217.2	3.0
7	1262.5	1222.4	3.3
8	1270.4	1224.4	3.8
9	1274.9	1223.3	4.2
10	1281.7	1219.6	5.1
11	1285.2	1213.6	5.9
12	1287.8	1205.8	6.8
13	1294.7	1196.7	8.2
14	1285.1	1186.9	8.3
15	1230.4	1176.9	4.6