基于神经网络模型预测的水声对抗子母弹跌落冲击仿真

为了评估水声对抗子母弹在不同跌落冲击条件下的可靠性,应用LS-DYNA对母弹在不同跌落参数下的冲击响应进行了数值仿真研究。在有限元计算结果作为样本的条件下,运用BP神经网络模型建立了母弹跌落冲击影响参数与子弹应力峰值的非线性映射关系。结果表明:弹体各子结构在跌落高度、角度及跌落面条件下具有不同的冲击响应规律;母弹体头螺、壳体结构冲击变形较大,内部子弹冲击应力值变化幅度较小;网络模型较准确预测子弹的应力峰值,可极大地降低仿真次数,提高跌落测试效率。

Drop impact simulation of the underwater acoustic countermeasure cluster ammunition based on the prediction model of neural network

In order to assess the reliability of the underwater acoustic countermeasure cluster ammunition in different drop impact condition, numerical simulation of shock response of the dispenser in different drop parameters was studied by LS-DYNA. The nonlinear mapping relation between the drop impact parameters of dispenser and the stress peak of the bullets was established by BP neural network model. The results showed that the rule of impact response of each substructure of the projectile body was different in different drop height, angle and drop surface. The deformation of head screw and shell structure of the projectile was relatively large, the change range of stress value of impact in bullets were relatively small, and the stress peak of bullets was predicted relatively accurately by network model, simulation times can be reduced greatly, and efficiency of drop test also can be improved simultaneously.