ANN在一次引爆FAE爆炸威力评价中的应用

在系统分析燃料空气炸药(FAE)爆炸威力主要影响因素的基础上，应用人工神经网络(ANN)理论及方法，建立了一次引爆FAE爆炸威力评价ANN模型，并利用该模型对不同配比环氧丙烷／铝粉混合燃料的爆轰参数及爆炸威力进行了模拟。结果表明，ANN模拟结果与一维多相爆轰模型计算值相当吻合，当铝粉含量处于50％～70％范围时，爆炸威力较为理想，当铝粉含量为65％时，爆炸威力指数达到最大。这一研究可为一次引爆FAE配方及战斗部设计与优化提供有益参考。

Application of ANN in Explosion Power Evaluation for Single Ignition FAE

Based on a systematic analysis of primary affected factors of fuel air explosive (FAE) explosion power and on the application of artificial neural network (ANN) theory and method,an ANN model is constructed for explosion power evaluation of single ignition FAE,detonation parameters and explosion power of fuel mixtures containing different ratios of liquid propylene oxide and solid aluminum powder are also simulated with the constructed ANN model.The results show that simulation of the ANN model is well consistent with the calculation value of one dimension models for multi-phase detonation.Explosion power index will reach its maximum if the content of aluminum powder is 65% and better explosion power will be obtained if the content of aluminum powder is between 50% and 70%.The study can provide helpful references for designing and optimization of the single ignition FAE batch formula and warhead.