Aimed at current deficiencies of multi-purpose guided missile kill probability model against gunship, the concept of the important coefficient of vulnerability blade unit is proposed in this paper. Laser fuze actuatio...Aimed at current deficiencies of multi-purpose guided missile kill probability model against gunship, the concept of the important coefficient of vulnerability blade unit is proposed in this paper. Laser fuze actuation model and warhead condition kill probability model of rotor blades are established by Monte Carlo method and kinetics theory with new ideas. Based on limited data, armor thickness of gunship is estimated, and a complete multi-purpose guided missile kill probability mathematical model is established, which provides necessary mathematical tool for the accurate and objective analysis of multi-purpose guided missile kill probability against gunship. Based on the establishment of the model, sensitivity analysis and optimal design of the main factors of multi-purpose guided missile kill probability are conducted, and the results show that the single multi-purpose guided missile lethality performance can be improved significantly by sensitivity analysis and optimization.展开更多
文摘机器学习模型(Machine Learning,ML)的不可解释性给其在气象业务中的应用带来了挑战。模型解释和可视化是解决这一问题的有效途径。文中将SHAP值应用于天气预报ML模型解释,研究了江西省暖季暴雨模型的预报因子对预报结果的影响。分别选取2016—2020年、2021—2022年4—9月ECWMF(European Centre for Medium-Range Weather Forecasts)高分辨率数值模式物理量及国家站降水观测数据进行XGBoost建模与模型解释。结果表明,全局重要性排名前4位依次是总降水(重要性42.70%)、850 hPa比湿(重要性11.17%)、925 hPa相对湿度(重要性10.44%)、500 hPa相对湿度(重要性9.16%)。个例分析表明,命中个例中高重要性物理因子在暴雨区的SHAP值较大,漏报(空报)个例在漏报(空报)区域高重要性物理因子的SHAP值偏小(偏大)。SHAP值从全局和局部可定量给出ML模型有物理意义的解释,解释结果与天气学原理和业务经验较一致,有利于ML在气象业务中的深入应用。
文摘采用顶空固相微萃取(headspace solid-phase microextraction,HS-SPME)结合气相色谱-质谱联用仪(gas chromatography-mass spectrometry,GC-MS)分别对普洱市5个产茶区的普洱生茶香气组分进行分析。结果表明,42个普洱生茶样品中共检测出83种香气成分,其中醇类化合物23种,碳氢类化合物20种,酯类化合物12种,甲氧基苯类化合物8种,酮类化合物8种,酸类化合物4种,酚类化合物3种,醛类化合物3种,含氮类化合物2种。不同产茶区普洱生茶中醇类、碳氢类和醛类化合物相对百分含量差异均不显著,甲氧基苯类化合物含量差异显著性比例较大,酯类、酮类、酚类、酸类和含氮类化合物含量差异显著性比例较小。以83种香气成分相对百分含量为变量进行偏最小二乘判别分析(partial least squares-discriminant analysis,PLS-DA),不同产茶区的普洱生茶样品呈现明显的分离趋势,变量重要性因子(variable importance in the projection,VIP)分析,33种香气成分对不同产茶区普洱生茶样品区分起主要作用(VIP> 1),33种香气成分中甲氧基苯类、酮类、酯类和醇类物质占比最大。
文摘Aimed at current deficiencies of multi-purpose guided missile kill probability model against gunship, the concept of the important coefficient of vulnerability blade unit is proposed in this paper. Laser fuze actuation model and warhead condition kill probability model of rotor blades are established by Monte Carlo method and kinetics theory with new ideas. Based on limited data, armor thickness of gunship is estimated, and a complete multi-purpose guided missile kill probability mathematical model is established, which provides necessary mathematical tool for the accurate and objective analysis of multi-purpose guided missile kill probability against gunship. Based on the establishment of the model, sensitivity analysis and optimal design of the main factors of multi-purpose guided missile kill probability are conducted, and the results show that the single multi-purpose guided missile lethality performance can be improved significantly by sensitivity analysis and optimization.
