填充式防护结构弹道极限方程的多指标寻优

为获得适用于国内填充式防护结构超高速撞击的弹道极限方程，采用多指标寻优的方法，对NASA填充式防护结构的弹道极限方程以国内实验数据为依据进行修正.结果发现：采用第1类指标（总体预测率和安全预测率）和第2类指标（预测误差平方和）联合对方程的系数进行修正，可获得预测效果更好的修正方程.通过对方程低速段和高速段的整体系数进行修正，最终获得单填充组、单一材料的双填充组以及两种材料的双填充组防护结构弹道极限方程的总体预测率分别为93.3%，90%和88.9%，而安全预测率全部高达100%，可很好满足工程的需求.可见，基于不同填充式防护结构的实验数据分别进行弹道极限方程的修正，可获得相应结构预测能力较优的方程.

Multiple indicators optimization for stuffed Whipple shield ballistic limit equation

In order to obtain the hypervelocity impact ballistic limit equation suiting for domestic stuffed Whipple shield, multiple indicators optimization method was studied. Based on domestic data, the NASA's ballistic limit equation of stuffed Whipple shield was revised. The results show that, joint the first kind of indicators(totality predicted rate and safety predicted rate) and the second kind of indicator (sum of squared prediction errors) together, better predicted equation can be obtained by correcting the coefficients of the stuffed Whipple shield ballistic limit equation. By correcting the coefficients of ballistic region and hypervelocity region, the corrected ballistic limit equations-totality predicted rate achieve 93.3%, 90% and 88.9% respectively for single-filled group, dual-filled group with single material and dual-filled group with two materials. And their safety predicted rate all reach up to 100%. So they can perfectly meet the engineering requirements. It is obvious that correcting the ballistic limit equations basing on experimental data for different kinds of stuffed Whipple shields, better predicted equations can be obtained for corresponding structures.