欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2024, Vol. 45 ›› Issue (6): 1889-1905.doi: 10.12382/bgxb.2023.0094

• • 上一篇    下一篇

基于人工神经网络算法的多相云雾爆轰毁伤效应预测模型

徐永康, 薛琨*()   

  1. 北京理工大学 爆炸科学与技术国家重点实验室, 北京 100081
  • 收稿日期:2023-02-16 上线日期:2023-06-06
  • 通讯作者:

Artificial Neural Network-based Prediction Model for Damage Effect of Fuel-air Explosive

XU Yongkang, XUE Kun*()   

  1. State Key Laboratory of Explosion Science and Technology,Beijing Institute of Technology,Beijing 100081,China
  • Received:2023-02-16 Online:2023-06-06

摘要:

云雾爆轰毁伤范围的预测是云爆武器造成大规模毁伤的研究基础,但云雾爆轰后超压场分布规律与燃料浓度的依赖关系未知,制约了对于云爆毁伤范围的预测。因此,针对2种最为常用的多相云雾燃料,采用最小自由能法计算得到了液相燃料完全以液滴或蒸汽形式存在的多相云雾发生理想爆轰的CJ参数,并通过拟合,得到体JWL状态方程参数。在此基础上计算得到了不同浓度和状态的多相云雾理想爆轰造成的超压场,并借助人工神经网络获得了浓度范围在0.03~0.30kg/m3的气固两相和气液固三相云雾场爆轰峰值超压随比例距离衰减规律的代理模型并预测不同毁伤等级对应的毁伤比例半径随燃料浓度的变化,得到毁伤比例半径最大的最优浓度。研究结果表明:云雾区中液相燃料以液滴或蒸汽形式存在对云雾爆轰参数,产物JWL状态方程参数,与云爆爆轰后超压场分布规律的影响都比较微弱(<1.5%);在0.03~0.18kg/m3的燃料浓度范围内,Ⅰ级~Ⅲ级毁伤比例半径的最大和最小值分别相差21%、19%、6%,因此大装药结构形成的云雾场爆轰后,Ⅰ级和Ⅱ级毁伤半径与燃料浓度的依赖性更强。

关键词: 云雾爆轰, 燃料浓度, 超压毁伤, 人工神经网络

Abstract:

The prediction of damage range caused by Fuel-air Explosive is fundamental to the study of large-scale damage caused by Fuel-air Explosive weapons. However, the distribution pattern of shock waves after detonation and its dependence on fuel concentration are unknown, which limits the prediction accuracy of damage range. In this study, the minimum free energy method is used to calculate the CJ parameters for the ideal detonation of biphasic cloud fog with liquid fuel present in either droplet or vapor form. The JWL equation of state parameters are obtained through fitting. Subsequently, the peak overpressure caused by ideal detonation of biphasic cloud fog with different concentrations and states is calculated. A proxy model is developed by utilizing an artificial neural network. The proposed model is used to predict the decay law of peak overpressure with respect to the scaled distance for biphasic gas-solid and gas-liquid-solid cloud detonations with concentrations ranging from 0.03 to 0.30kg/m3. The model is also used to predict the variation of damage proportion radius with fuel concentration for different damage levels, obtaining the optimal concentration with the maximum damage proportion radius. The study reveals that the influences of liquid fuel in droplet or vapor form on cloud detonation parameters, JWL equation of state parameters, and shock wave distribution after cloud detonation are relatively weak (<1.5%). Within the fuel concentrations ranging from 0.03 to 0.18kg/m3, the maximum and minimum values of damage proportion radius for damage levels Ⅰ-Ⅲ are differed by 21%, 19%, and 6%, respectively. Thus, the dependence of damage radii on fuel concentration is stronger for damage levels Ⅰ and Ⅱ after cloud burst caused by large explosive structures.

Key words: cloud detonation, fuel concentration, overpressure damage, artificial neural network

中图分类号: