基于双Markov链的效果评估模型研究  

Study of Double-Markov Chains to Estimating Effect of Naval Mine-sweeping

作　　者：李庆民[1] 王红卫[1] 李华[2] 刘君[2] 

机构地区：[1]华中科技大学控制科学与工程系,武汉430074 [2]海军工程大学兵器工程系,武汉430033

出　　处：《武汉理工大学学报（交通科学与工程版）》2007年第3期460-463,共4页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国防预研项目资助

摘　　要：当前基于单Markov链的扫雷作战效果评估模型是一个理论意义大于实际的模型,如应用于实际的多Markov扫雷过程则还有相当的距离.文中从正交分解的角度,指出Markov评估模型只能得到定性评估结果.进一步提出了用于定性评估的双Markov模型,计算结果表明:在引入先验知识后,能在小样本数据条件下取得较为准确的评估结果.For the multivariate Markov chains existing in the process of naval mine-sweeping, there is a long way to be applied in fact for the effect estimating model based on single Markov chain. From the view of orthogonal factorization, the single Markov chains estimating effect model can only get the blur outline of mines. Double Markov chains model has been set up to estimate effect of naval minesweeping in the paper. Simulation result shows that the estimating model of double Markov chains can get more exact outline of mines with the help of experience knowledge even in the condition of small sample data.

关 键 词：MARKOV链 遗传算法 扫雷作战效果 正交基 

分 类 号：TJ01[兵器科学与技术—兵器发射理论与技术]