基于数据融合理论的多传感器几何定位算法

Geometric location algorithm for multi-sensor networking based on data fusion theory

为解决多部(3部及以上)2D传感器网络对三维空间目标的定位估计和定位精度问题,克服地球曲率对观测模型的影响,建立了考虑实际地球曲率的等效地球模型和传感器观测模型,提出了此模型中基于二次数据融合的多传感器组网几何定位算法,该方法将几何定位与数据融合理论相结合,并对融合数据进行二次融合,充分利用了各传感器的量测数据。仿真实验证明了方法的有效性和实用性,在多部2D传感器组网的情况下可对三维空间内目标实现精确定位,定位误差趋近于克拉美-罗下界(CRLB),具有工程实用价值。

To solve the problems of estimation and accuracy for target location in 2D sensor network with multi-sensors, the actual models of equivalent earth and geometric sensor observation were built considering the influence of the actual earth curvature on the observation model. In order to make full use of the measurements, a geometric location algorithm for multi-sensor networking based on 2D data fusion was presented, and the location equations were calculated. This algorithm combined the methods of geometric location and data fusion, and carded out two times of data fusion to utilize all the measurements. In order to analyze the performance of the proposed algorithm, some computing cases were illustrated. The efficiency and superiority of the proposed algorithm can be confirmed by simulation results. It can be seen from the results that the algorithm can improve locating accuracy greatly in 3D space, and the accuracy tends to Cramer-Rao lower bound（CRLB）. Although there are two times of data fusion, it does not increase the complexity of calculation, and the proposed algorithm has practical value and validity.