基于随机森林的无源时差定位误差估计与误差修正算法  

Estimation and Correction Algorithm for Passive TDOA Localization Based on Random Forest

作　　者：宋定宇 张君毅[1,2] SONG Dingyu;ZHANG Junyi(The 54th Research Institute of China Electronics Technology Group Corporation,Shijiazhuang 050081,China;Hebei Province Key Laboratory of Electromagnetic Spectrum Cognition and Control,Shijiazhuang 050081,China)

机构地区：[1]中国电子科技集团公司第54研究所,石家庄050081 [2]河北省电磁频谱认知与管控重点实验室,石家庄050081

出　　处：《计算机测量与控制》2025年第1期269-275,284,共8页Computer Measurement &Control

基　　金：国家自然科学基金(U20B2071)。

摘　　要：针对无源时差定位算法在时差估计误差未知条件下,无法通过计算几何稀释因子准确估计定位误差的问题,提出基于随机森林的无源时差定位误差估计与误差修正算法,通过学习信号特征参数和定位信息与定位误差之间的映射关系,准确估计定位误差并对定位结果进行修正,实现高精度定位;通过匹配时差定位结果与参考源信息制作了测试数据集,并使用该数据集验证了所提算法的有效性;对各特征参数的重要性进行了定量分析,并验证了随机森林模型在时间上的泛化能力;实验结果表明,所提算法实现了在时差估计误差未知条件下对时差定位误差的准确估计,提高无源时差定位的精度,具有较高的工程应用价值。Under the condition of unknown TDOA estimation error,passive time difference of arrival(TDOA)positioning algorithm can not accurately estimate positioning error by calculating geometric dilution of precision(GDOP),this paper proposes a TDOA positioning error estimation and correction algorithm based on random forest.By learning the mapping relationship between signal feature parameters,positioning information,and positioning error,it accurately estimates the positioning error and corrects the positioning result to achieve high-precision positioning.A test dataset was created by matching the TDOA positioning result with reference source information,and the effectiveness of the proposed algorithm is verified by this dataset.The importance of each feature parameter is quantitatively analyzed,and the temporal generalization ability of the random forest model is validated.Experimental results show that the proposed algorithm achieves accurate estimation of TDOA positioning error under the condition of unknown TDOA estimation error,improves the accuracy of passive TDOA positioning,and has high engineering application value.

关 键 词：无源定位 时差定位 随机森林 误差估计 误差修正 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]