针对精确模型研究弹药温度变化的计算精度低、可靠性差的问题,提出了运用BP神经网络研究弹药传热规律的方法,建立了基于BP神经网络的弹药传热模型,以某两型榴弹传热实测数据对模型进行了训练,利用训练后模型对另一型榴弹的传热规律进行了仿真计算,结果表明:该模型逼近精度高、计算速度较快,可用于预测弹药的传热规律。

For low calculation accuracy and reliability problem of accurate model in studying ammunition temperature change, BP neural network method is proposed to study the law of heat transfer of ammunition, and the ammunition heat transfer model based on BP neural network is established. The model is trained with the measured heat transfer data of two types of ammunitions, and then the model is used to simulate the heat transfer law of another type of ammunition. The results show that the model owns high approximation precision and fast calculation speed, and can be used to predict the heat transfer law of ammunitions.