摘要
为降低强光爆震弹在使用中潜在的安全威胁,以非致命效应为出发点,采用均匀设计法进行了配方设计,基于BP神经网络建立了装药性能预测模型,采用神经网络与遗传算法相结合的方法进行了装药优化.通过声光效应试验对优化结果进行了验证,得到了最优配方:KClO4/Al/CS/环氧树脂/石墨=48/32/15/2/3.结果表明,通过装药配方的优化,提高了闪光爆震弹的综合性能,为防暴弹的性能改进提供了新的思路和方法.
In order to reduce the potential security threat to living targets,this paper takes the non-lethal effects as a starting point,adopts the uniform design method for the charging design,establishes a charge performance prediction model based on BP neural network,and optimizes the charge by combining neural network with genetic algorithm approach.The optimal results are verified through acousto-optic effect test,and the optimal recipe is obtained：KClO4/Al/CS/Epoxy Resin/Graphite=48/32/15/2/3.Results show that the optimization of the charging formula improves the overall performance of the flashing detonation grenades,providing new ideas of the formulation design and performance improvement methods for the research of riot bombs.
出处
《军械工程学院学报》
2016年第4期27-31,共5页
Journal of Ordnance Engineering College
关键词
强光爆震弹
均匀设计法
遗传算法
BP神经网络
flashing stun grenade
the uniform design method
genetic algorithm approach
BP neural network
