中文摘要：

传统的多站角度定位系统需要各站先估计目标到达方向角（DOA）,然后估计目标位置,在低信噪比时往往会导致角度估计误差太大而无法定位。提出了一种单目标多阵列子空间数据融合（SDF）直接定位算法。该算法首先通过传统方法对各阵列输出数据进行自相关运算和特征值分解,得到各站的噪声子空间数据,再在监测区域内用量子粒子群算法搜索在各噪声子空间投影的和最小的位置矢量,估计目标位置。仿真实验表明,这种算法在低信噪比下较传统的交叉定位算法有更好的性能。

英文摘要：

Traditional multi-station localization systems need to estimate the direction of arrival （DOA） to determine the location of the target. In low SNR situations the localization fails because of the large estimation errors. In this paper, a direct localization algorithm for a single target with multi-array using subspace data fusion （SDF） is proposed. The output data of each array is first pro- cessed with self-correlation and eigenvalue decomposition to get the noise subspace data. The posi- tion of the target is then determined by searching the best array response in the monitoring area which has the biggest sum of projections to respective noise subspace using quantum-behaved parti- cle swarm optimization （QPSO） algorithm. Simulation results demonstrate that it performs better in the low SNR situation than the bearing-crosslng localization algorithm.