修正容积卡尔曼滤波数据域直接定位算法

为解决数据域直接定位（DPD）算法面临的计算压力，提高算法效率，提出一种基于修正的容积卡尔曼滤波（MCKF）的DPD算法。首先，融合各观测信号波达方向信息，利用子空间数据融合方法建立一种基于间接观测量的DPD滤波模型；然后，根据模型特点设计MCKF算法进行求解，解决间接观测量带来的噪声累积问题；最后，对算法计算量进行分析和对比，说明计算效率的提升。仿真结果表明，所提算法与基于最大似然遍历搜索和遗传算法的DPD算法相比，在相同的估计性能下，计算量下降明显，时效性显著提升，增加了算法实用价值。

Direct position determination algorithm based on modified cubature Kalman filter

The Direct Position Determination (DPD) algorithm can offer higher estimation accuracy, but is accompanied by serious computation load. To solve this problem, this paper proposes a novel DPD algorithm based on Modified Cubature Kalman Filter (MCKF). First, a direct positioning filtering model based on the direction of arrival information of the received signals is developed with the subspace data fusion method. Then, an MCKF is proposed to solve the problem of noise accumulation caused by indirect observation. Finally, the computational complexity of the algorithm is analyzed to illustrate its efficiency. Simulation results show that compared with the DPD algorithms based on maximum likelihood exhaustive search and genetic algorithm, the proposed algorithm can significantly reduce computation load and improve timeliness while maintaining the same estimation accuracy, showing great applicability in practice.