摘要
采用楔形装药,测量了粒径范围在1~2μm（平均粒径为1.5μm）的超细黑索今（RDX）在不同密度、粒度与不同粘结剂含量下的爆轰临界厚度。以密度、粒度与粘结剂含量作为输入元,RDX爆轰临界厚度作为输出元建立了RDX爆轰临界厚度的E lman神经网络预测模型。以实验数据对模型进行训练,采用达到误差要求的模型对不同密度、粒度与粘结剂含量下的RDX爆轰临界厚度进行预测。预测结果显示：3个影响因素对临界直径的影响规律与文献[9-11]报道相同,说明神经网络用于RDX爆轰临界厚度预测是可行的。
The critical thickness of RDX at different densities,grain sizes and binding content was tested using a wedge-shaped charge.The Elman prediction model was established using density,grain size and binding content as input variables,and the critical thickness of RDX as output variable to predict the critical thickness under other conditions.The results show that the relationship between the three factors and the critical thickness is the same as that in Ref.,and the neural network can be used to predict the critical thickness of RDX.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2010年第10期1394-1397,共4页
Acta Armamentarii
