ID	原文	译文
10864	所提算法计算量当目标相距较远时低于MM-PHD和MM-CBMeMBer,目标聚集时增长速度快于对比算法。	roposed algorithm calculation when the target is far below the MM apart - PHD and MM - CBMeMBer, goal gather growing faster than comparison algorithm. ‭
10865	为拓展反导作战中红外预警卫星的目标速度估计方式,提出利用星载探测器进行被动红外多普勒频移测速的构想。	To expand the anti-missile combat of early warning satellite infrared target velocity estimation, put forward the use of satellite-borne detector for passive infrared doppler frequency shift speed.
10866	首先建立多星多普勒频移测速模型,给出了多普勒频移测速误差与观测几何、辐射波长以及光谱分辨率的关系;	First establish multiple doppler frequency shift velocity model, and gives the doppler frequency shift speed measuring error and observation geometry, the relationship between radiation wavelength and spectral resolution; ‭
10867	‭然后推导多普勒频移测速误差与导弹发射坐标系下速度误差的坐标转换关系,描述了多普勒频移测速对弹道参数估计的贡献。	Then deduced doppler frequency shift speed error and velocity error of coordinate transformation relations under the missile launch coordinate system, describes the doppler frequency shift velocity of ballistic parameters estimation of contribution.
10868	最后,结合战术需求,通过仿真实例分析了多普勒频移测速对参数估计的影响。	Finally, combining with the tactical demand, doppler frequency shift was analyzed through simulation examples for parameter estimation of speed.
10869	针对多分类极限学习机(extreme learning machine,ELM)缺乏概率输出能力问题,提出一种基于sigmoid后验概率映射和Lagrange成对耦合法的多分类概率ELM(multi-class probabilistic ELM,MPELM)。	Classification for more extreme learning machine (extreme learning machine, ELM) lack probability output ability problem, put forward based on the posterior probability mapping and sigmoid though laser coupling method in pairs more classification probability ELM (multi - class probabilistic ELM, MPELM).
10870	采用成对耦合法将多分类问题分解成多个二分类问题,利用sigmoid函数将二分类ELM输出映射成概率输出。	‭Using coupling method in pairs to multiple classification problem is decomposed into multiple binary classification problems, sigmoid function is used to change the binary classification ELM output mapping probability output.
10871	为融合所有二分类概率输出,推导基于Lagrange乘子法的多分类概率计算公式,最终求解被预测样本分属不同类别的概率。	‭All binary classification to fusion probability output of multiple classification based on though laser multiplier method is derived formula of probability, the final solution was predicted samples belong to different categories of probability.
10872	将MPELM用于剩余使用寿命(remaining useful life,RUL)预测,实验结果表明,相比于多分类概率支持向量机(multi-class probabilistic support vector machine,MPSVM),MPELM耗时高于MPSVM,但MPELM所需优化参数少,预测精度高于MPSVM;	‭MPELM for remaining service life (remaining useful life, url) prediction, the experimental results show that compared with multiple support vector machine (SVM) classification probability (multi - class probabilistic support vector machine, MPSVM), MPELM time-consuming than MPSVM, but MPELM less needed to optimize parameters, prediction accuracy is higher than MPSVM;
10873	与基于Hastie成对耦合法的MPELM相比,两者预测精度相近,本文MPELM的测试耗时较少。	Compared with MPELM based on Hastie pair coupling method, the prediction accuracy, this paper MPELM test less time-consuming.