基于马尔科夫链的交互残差估计的跟踪波门设计方法

跟踪波门通常基于残差估计进行设计,但传统残差估计法在目标机动状态未知时的预测精度较差。针对这个问题,在分析2种现有残差估计法的适用环境基础上,提出了一种基于马尔科夫链的交互残差估计法:该方法首先分别利用现有2种估计法对下一时刻残差值进行预测,定义并计算滑窗内2种方法的预测误差,并根据预测误差的似然函数计算对应似然值;然后,基于服从已知概率的马尔科夫链的转移矩阵,采用似然值求解各方法所占权重;最后,通过加权求和得出下一时刻残差预测值。仿真实验表明,利用文中算法得出的残差预测值在目标运动状态未知的情况下精度较高,据此设计的跟踪波门性能更优。

A Tracking Gate Design Method of the Interaction Estimation Based on Markov Chain

Design of gate tracking is usually made based on residual estimation, but the traditional residual estimation method is comparatively poor in precision under condition of the unknown target state. For this problem, this paper analyzes the application environments of two kinds of the existing residual estimation methods, and proposes an interactive estimation algorithm based on Markov chain. Firstly, the existing two kinds of estimation methods are used to predict respectively the residual value of next time. The prediction error is defined and calculated according to the predicted values. Then, the likelihood value is used to calculate the weight of each methods based on the transition matrix of Markov chain. Finally, the residual value of next time is sought out through weight summation. The simulation experiments show that the prediction accuracy obtained by this method is higher than that by the traditional methods under condition of the unknown target state.