蜂窝夹层板撞击极限方程预测能力的提升

蜂窝夹层板撞击极限方程是空间碎片撞击航天器风险评估的关键技术，目前描述其预测能力的指标主要有总体、安全预测正确率和绝对、相对误差。基于131个蜂窝夹层板的实验数据，分别描述各个预测指标在方程系数空间的变化特征，并采用层次化思路进行方程预测指标提升的探讨。结果发现，进行方程优化时，预测概率型指标可精确优化，而预测误差型指标可快速优化；总体预测正确率作为首要预测指标可优先用于研究航天器的在轨防护特性，而安全预测正确率作为首要指标则可优先用于其设计安全性。

Promotion of prediction ability of impact limit equationon honeycomb sandwich panel

Impact limit equations of honeycomb sandwich panel are the key of risk assessment of the spacecraft's impact from space debris. At present, the indicators of prediction ability mainly are the rate of correctly prediction (including the total rate and the safe rate) based on the prediction probability, and prediction error (including the absolute error and the relative error) based on the deviation of diameter predicted. In order to improve the prediction ability of equations, the variation of each predictive indicator in coefficient space of the equations was described based on 131 experimental data of honeycomb sandwich panel from the literature. And hierarchical analysis was used to investigate the promotion of prediction indicators of the equations. The results show that, equation optimization lead to accurately optimization of indicators of prediction probability, quickly optimization of prediction error. The total rate of correctly prediction can be used in the study of spacecraft's in-orbit protective properties as the primary indicator, and the safe rate of correctly prediction can be used in the security evaluation of design as the primary indicator. The prediction ability of the optimized equation has been greatly enhanced compared with the original equation.