针对无人机集群无源定位领域的目标跟踪问题,提出了一种基于改进到达时间差(TDOA)算法与卡尔曼预测模型的目标定位跟踪算法。首先,针对现有基于Chan的TDOA算法在时差值精度不够时定位精度差,以及现有基于Taylor的TDOA算法受迭代初始值影响大导致算法稳健性不足的问题,提出基于Chan-Taylor的改进TDOA算法,该算法通过Chan算法快速求解出目标位置初始估计值,然后使用Taylor算法迭代求解出目标位置,在提高定位效率、缩短定位时间的基础上,提高了定位精度。其次,利用卡尔曼预测模型,将目标运动模型产生的目标位置预测值作为先验信息带入目标定位算法,通过将预测信息与定位信息有效融合,进一步提高目标定位精度。最后,通过实验仿真,从累积分布函数(CDF)、定位精度几何因子(GDOP),以及定位跟踪RMSE多个方面验证文中所提算法的正确性及有效性。

In this article, a target positioning and tracking algorithm based on an improved time difference of arrival (TDOA) algorithm and a Kalman prediction model is proposed, addressing the target tracking problem in the field of passive localization for drone swarms. Firstly, addressing the shortcomings of the existing Chan-based TDOA algorithm in poor time difference precision leading to inaccurate positioning and the robustness issues of the Taylor-based TDOA algorithm due to the significant impact of iterative initial values, a Chan-Taylor combined TDOA algorithm is introduced. This algorithm rapidly determines the initial estimate of the target position using the Chan algorithm and then iteratively calculates the target position using the Taylor algorithm. Based on improving positioning efficiency and shortening positioning time, the positioning accuracy has been enhanced. Secondly, utilizing the Kalman prediction model, the predicted target position values generated by the target motion model are incorporated as prior information into the target positioning algorithm. By effectively integrating predictive and positional information, the precision of target positioning is further improved. Finally, through experimental simulations, the proposed algorithm′s correctness and effectiveness are validated from several aspects, including the cumulative distribution function (CDF), geometrical dilution precision(GDOP) of positioning accuracy, and the root mean square error (RMSE) of positioning tracking.

TN957;V279