基于神经网络的熔铸炸药热物性参数反演

蒋杰; 张向荣; 岳跃辉; 钟光天; 刘攀; 周霖

doi:10.12382/bgxb.2025.0797

您当前的位置：

首页 >

文章列表页 >

基于神经网络的熔铸炸药热物性参数反演

更新时间：2026-06-16

- 基于神经网络的熔铸炸药热物性参数反演
- Inversion of Thermophysical Properties of Melt-cast Explosives via Artificial Neural Networks
- 兵工学报 2026年47卷第5期页码：250797
- 作者机构：
  
  北京理工大学机电学院，北京 100081
  重庆红宇精密工业集团有限公司，重庆 402760
- 作者简介：
  
  通信作者邮箱：zhangxr@bit.edu.cn
- 基金信息：
- DOI：10.12382/bgxb.2025.0797
  中图分类号： TQ560.1
- 收稿：2025-09-01，
  
  网络首发：2026-04-09，
  
  纸质出版：2026-05
- 稿件说明：
移动端阅览
蒋杰, 张向荣, 岳跃辉, 等. 基于神经网络的熔铸炸药热物性参数反演[J]. 兵工学报, 2026,47(5):250797.

JIANG Jie, ZHANG Xiangrong, YUE Yuehui, et al. Inversion of Thermophysical Properties of Melt-cast Explosives via Artificial Neural Networks[J]. Acta Armamentarii, 2026, 47(5): 250797.
蒋杰, 张向荣, 岳跃辉, 等. 基于神经网络的熔铸炸药热物性参数反演[J]. 兵工学报, 2026,47(5):250797. DOI： 10.12382/bgxb.2025.0797.

JIANG Jie, ZHANG Xiangrong, YUE Yuehui, et al. Inversion of Thermophysical Properties of Melt-cast Explosives via Artificial Neural Networks[J]. Acta Armamentarii, 2026, 47(5): 250797. DOI： 10.12382/bgxb.2025.0797.

摘要

针对于传统热物性参数测定依赖于实验测量，耗时费力且难以获取连续温度函数的问题，本文提出一种基于贝叶斯正则化优化双层BP神经网络的反演方法，以精确获取熔铸炸药基体DNAN在凝固过程中的关键热物性参数（导热系数与比热容

）。

通过构建以径向多个测温点冷却时间差为输入、对应温度下热物性参数为输出的映射关系，利用有限元仿真生成大量训练样本，并结合实验数据对网络进行训练与验证。研究结果表明，该神经网络模型在反演导热系数和比热容时决定系数

达到0.994，反演结果与实测数据对比平均误差小于5%，具有很好的预测精度和良好的工程适用性，为热物性参数的获取提供了新途径。

Abstract

The conventional methods for determining the thermophysical properties predominantly rely on the experimental measurements，which are often time-consuming，labor-intensive，and inadequate for obtaining the continuous temperature-dependent functions. To address these limitations，this paper proposes an inversion method based on a two-layer Backpropagation（BP）neural network optimized with Bayesian regularization. This method is used to accurately acquire key thermophysical parameters（thermal conductivity and specific heat capacity）in the solidification process of melt-cast explosives. A mapping relationship is constructed using the cooling time differences from multiple radial temperature measurement points as inputs and the corresponding thermophysical properties at specific temperatures as outputs. A substantial set of training samples are generated via finite element simulation，and the network is subsequently trained and validated with experimental data. The results demonstrate that the neural network model achieves a coefficient of determination（

）of 0.994 for the inversion of both thermal conductivity and specific heat capacity. The average error between the inverted results and the measured data is less than 5%，indicating high predictive accuracy and strong engi

neering applicability. This method provides a new approach for acquiring the thermophysical parameters of melt-cast explosives.

关键词

Keywords

references

谢志毅,王冬磊,刘建波,等. 炸药熔铸过程数值模拟研究[J]. 太赫兹科学与电子信息学报, 2014, 12(4): 617-621.

XIE Z Y, WANG D L, LIU J B, et al. Numerical simulation of melt-cast explosive solidification process[J]. Journal of Terahertz Science and Electronic Information Technology, 2014, 12 (4):617-621.(in Chinese)

马松,袁俊明,刘玉存,等. DNAN炸药凝固过程的实验与数值模拟[J]. 含能材料, 2014, 22(2): 240-244.

MA S, YUAN J M, LIU Y C, et al. Experiment and numerical simulation of DNAN solidification process[J]. Chinese Journal of Energetic Materials, 2014, 22(2): 240-244. (in Chinese)

蒙君煚,张向荣,周霖. DNAN基熔铸炸药成型过程数值仿真[J]. 兵工学报, 2013, 34(7): 810-814.

MENG J J, ZHANG X R, ZHOU L. Simulation of solidification process for DNAN-based melt-cast explosives [J]. Acta Armamentarii, 2013, 34(7): 810-814. (in Chinese)

李敬明,田勇,张伟斌,等. 炸药熔铸过程缩孔和缩松的形成与预测[J]. 火炸药学报, 2011, 34(2): 17-20, 55.

LI J M, TIAN Y, ZHANG W B, et al. Formation and prediction of shrinkage hole and shrinkage porosity in explosive during casting process[J]. Chinese Journal of Explosives & Propellants, 2011, 34(2): 17-20, 55.(in Chinese)

奚同庚. 无机材料热物性学[M]. 上海:上海科学技术出版社, 1981.

XI T G. Thermophysical properties of inorganic materials[M]. Shanghai: Shanghai Scientific and Technical Publishers, 1981.(in Chinese)

CZEL B, GROF G, KISS L, et al. Simultaneous measurement of temperature dependent thermophysical properties [J]. Japanese Journal of Applied Physics, 211, 50(11): 11RE05.

YANG C Y. Determination of the temperature dependent thermophysical properties from temperature responses measured at medium's boundaries[J]. International Journal of Heat and Mass Transfer, 2000, 43(7): 1261-1270.

李昶亮. 基于BP神经网络的固体合金材料热物性快速预测研究[D]. 济南:山东大学, 2021.

LI C L. Study on rapid prediction of thermophysical properties of solid alloy materials based on BP neural network [D]. Jinan: Shandong University, 2021.(in Chinese)

孙金金,朱彤,吴家正. 神经网络法求解墙体传热系数关键输入变量的分析[J]. 新型建筑材料, 2006(12): 61-64.

SUN J J, ZHU T, WU J Z. Analysis of key input variables for solving wall heat transfer coefficient using neural networks[J]. New Building Materials, 2006(12): 61-64. (in Chinese)

LIU X, PENG W, GONG Z Q, et al. Temperature field inversion of heat-source systems via physics-informed neural networks[J]. Engineering Applications of Artificial Intelligence, 2022, 113: 104902.

JI H L, QI L L, LYU M X, et al. Improved bayesian optimization framework for inverse thermal conductivity based on transient plane source method [J]. Entropy ( Basel, Switzerland), 2023, 25(4): 575.

BECK J V, LITKOUHI B, CLAIR JR R S C. Efficient sequential solution of the nonlinear inverse heat conduction problem [J]. Numerical Heat Transfer Part A-Applications, 1982, 5(3): 275-286.

DA SILVA-SANTOS C H, GOULART P R, BERTELLI F, et al. Nature inspired algorithms for the solution of inverse heat transfer problems applied to distinct unsteady heat flux orientations in cylindrical castings [J]. Journal of Intelligent Manufacturing, 2023, 34(5): 2407-2430.

SU H J, KANG J Y, LI Y, et al. Parallel bayesian optimization of thermophysical properties of low thermal conductivity materials using the transient plane source method in the body-fitted coordinate[J]. Entropy (Basel, Switzerland), 2024, 26(12):1117.

鲁得浦,王成恩. 基于人工智能和有限元方法的传热学反问题[J]. 工业技术创新, 2023, 10(2): 32-42.

LU D P, WANG C E. Inverse heat transfer problem based on artificial intelligence and finite element algorithm[J]. Industrial Technology Innovation, 2023, 10(2): 32-42.(in Chinese)

陈亚琴,彭浩,冯诗愚. 基于人工神经网络的微重力流动冷凝换热预测[J]. 南京航空航天大学学报, 2021, 53(6):989-995.

CHEN Y Q, PENG H, FENG S Y. Prediction of flow condensation heat transfer under microgravity based on artificial neural network[J]. Journal of Nanjing University of Aeronautics& Astronautics, 2021, 53(6): 989-995.(in Chinese)

BALAZS C, WOODBURY K A, GYULA G, et al. Simultaneous estimation of temperature-dependent volumetric heat capacity and thermal conductivity functions via neural networks [J]. International Journal of Heat and Mass Transfer, 2014, 68: 1-13.

刘午扬,殷长春,苏扬,等. 基于人工神经网络正演模拟的航空电磁数据变维数贝叶斯反演[J]. 地球物理学报, 2025, 68(8): 2928-2940.

LIUWY,YIN C C, SU Y, et al. Trans-dimensional Bayesian inversion of airborne electromagnetic data based on artificial neural network [J]. Chinese Journal of Geophysics, 2025, 68(8): 2928-2940.(in Chinese)

KROGH A. What are artificial neural networks? [J]. Nature Biotechnology, 2008, 26(2): 195-197.

康靓,米晓希,王海莲,等. 人工神经网络在材料科学中的研究进展[J]. 材料导报, 2020, 34(21): 21172-21179.

KANG L, MI X X, WANG H L, et al. Research progress of artificial neural networks in material science [J]. Materials Review, 2020, 34(21): 21172-21179. (in Chinese)

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 J G, WANG Q, LI J Q, et al. Fragment initial velocity prediction model of triangular prism directional warhead structure based on artificial neural network [J]. Acta Armamentarii, 2025, 46(3): 240346. (in Chinese)

李航宇,王楠楠,朱明瑞,等. 神经结构搜索的研究进展综述[J]. 软件学报, 2022, 33(1): 129-149.

LIHY, WANG N N, ZHU M R, et al. Recent advances in neural architecture search: a survey[J]. Journal of Software, 2022, 33(1): 129-149. (in Chinese)

杨秀芬.DNAN基熔注炸药热物性参数实验及反演分析[D]. 北京:北京理工大学, 2021.

YANG X F. Experimental and inversion analysis of thermophysical properties of DNAN-based explosive [D]. Beijing: Beijing Institute of Technology, 2021.(in Chinese)

浏览量

下载量

CNKI被引量

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于多小波基融合的脉冲激光抗烟雾干扰算法

基于高压过冷的熔铸炸药快速成型机理

基于神经网络的电静液起竖系统换级碰撞补偿积分鲁棒控制

基于多特征融合网络的空袭目标运动状态辨识

基于双层多目标优化的翼身融合水下滑翔机设计