基于神经网络模型的燃料空气混合物爆炸威力预测研究

摘要

The research about accurate prediction on explosion power of fuel air mixture is a difficult problem in ac-ademic circle.Aiming at the effective prediction problem on explosion power of fuel air mixture, the neural network method was used to design the multi-layer neural network model, and the practical prediction application was con-ducted.The results showed that the applied prediction method is simple and feasible, and can provide a new way for the prediction on explosion power of fuel air mixture.Compared with the 3-layer BP model, the prediction model designed in the paper can reduce training number, shorten training time, increase prediction accuracy, and the ap-plication advantage is more obvious.%燃料空气混合物爆炸威力准确预测研究是学术界的一个难题。针对燃料空气混合物爆炸威力有效预测问题，采用神经网络方法，设计多层神经网络模型，进行实际预测应用。应用结果表明，采用的预测方法简便、可行，可以为燃料空气混合物爆炸威力预测提供一种新途径。相比3层BP模型，设计的预测模型可以减少训练次数，缩短训练时间，提高预测正确率，应用优势较明显。