天基平台弹道导弹参数似然估计算法

Research about Space-based Ballistic Missile Parameters Estimation with Maximum Likelihood Algorithm

针对空天防御体系中弹道导弹的跟踪提出了参数最大似然估计概率假设密度(Probability Hypothesis Density,PHD)多目标跟踪算法。该算法将未知的、影响跟踪精度的关键因素从目标状态变量中分离出来,设计了针对该参数估计的最大似然估计器,同时为解决似然比整体过小无效的问题,提出了利用马氏距离平方根计算似然比的方法。此外考虑弹道导弹实时跟踪的需求,推导并引入了新型积分点采样准则,以较少的Sigma点去近似高斯函数积分。仿真结果验证了改进算法的性能,可为反导拦截提供更为准确的情报信息。

Aiming at the Ballistic Missile Tracking in air-space defense system,a novel PHD algorithm is proposed.The novel PHD algorithm takes away critical unknown parameter from state,and certain parameter is provided with a Maximum Likelihood Estimator.Mean while,the method of calculating the tiny likelihood ratio with the square root of Mahalanobis distance is given in order to guarantee the efficiency of particles.Besides,considering the requirement of real-time Ballistic Missile Tracking,the article brought in and deduced a novel scaled sigma points.Results have proved that the novel algorithm is effective,and can provide more accurate information for missile defense and interception.