摘要
传统的燃烧模型致力于对固体推进剂燃烧的物理化学过程的数学描述，其发展往往受到对推进剂燃烧机理认识程度的限制，本文利用误差反传（ＢＰ）神经网络建模的方法建立了硝胺固体推进剂燃烧性能及其影响因素（硝胺含量、压强、硝胺重均粒径）的定量关系，而不考虑燃烧的具体过程，根据此定量关系对硝胺固体推进剂燃烧性能的计算表明，计算结果和实验值吻合得较好，这为推进剂配方的计算机辅助设计提供了一种新方法。
Traditional model of composite propellant combustion,which strives to represent the combustionprocess in mathematical terms,is restricted by the knowledge of the combustion mechanism.A new scheme,is proposed to calculate the combustion of composite propellant without considering the specific combustionprocesses by applying the back-propagation(BP)neural network to find the quantitative relationships between the factors(pressure,weight,average diameter of HMX,mass concentration of HMX)and combustion characteristics.According to the quantitative relationships, the combustion characteristics of HMX propellant are calculated,and the results are in good agreement with the experimental date.The scheme offers anew way for CAD of propellant formulation.
出处
《固体火箭技术》
EI
CAS
CSCD
1996年第3期17-22,共6页
Journal of Solid Rocket Technology
关键词
复合推进剂
推进剂燃烧
奥克托今
神经网络
Network Composite propellant Speculation Propellant combustion Octogen
HMX
