火炮发射药装药自动识别方法

Identification Technology Research for Powder Model of Artillery

发射装药号是影响火炮身管寿命的重要因素,为自动识别火炮发射药的装药号,避免人工记录失误和完善火炮射弹履历,分析了装药号自动识别原理,运用火炮动力学分析理论和ADAMS虚拟样机技术,建立火炮发射动力学虚拟样机。通过仿真试验获取样本数据,应用BP神经网络进行学习和训练,从而建立装药号和测试数据之间的非线性映射关系,实现对火炮发射装药号的精确预测。

In order to identify the powder model of artillery, which is the important factor to affect the life of cannon barrel, avoiding the manually record mistake and supplementing the firing record, the thesis focusing this theme includes the following work： analyzsis of the identify theory of charge, establish ment of the virtual prototype of a type of howitzer based on the dynamics analysis technology and ADAMS software. The sample couple is obtained by simulation. The back propagation algorithm is used to study and train the sample couple based on artificial neural network, and then the nonlinearity relation is achieved between the powder model and the influencing parameters. Therefore, it can provide the accurate measure to forecast the cannon charge.