纯方位单站目标被动跟踪需要观测站作机动才能满足可观测条件,导致跟踪算法收敛时间较长。基于纯方位多站被动跟踪算法可以解决算法收敛时间较长的问题,文中提出了一种约束最小二乘算法用于纯方位多站被动跟踪,它首先为最小二乘算法引入约束条件,利用矩阵最小特征值所对应的特征向量,解决了EKF 算法需要初值的问题,避免了滤波的发散,同时也极大地减小了最小二乘估计的偏置。仿真表明,这种算法能够渐进地逼近估计误差的下限,并且其精度有所增加,是一种渐进、稳定以及近似无偏的估计算法。

The sensor needs to maneuver to get better observability in bearings-only passive target tracking with single sensor which makes the observation time longer. Multisensor bearings-only passive target tracking can solve the problem using exchange data. So the constrained least square estimation(CLSE) algorithm is proposed for multisensor bearings-only passive target tracking in this paper. The constrained condition is introduced to the least square estimation algorithm firstly. Then the eigenvector corresponding to the least eigenvalue of the matrix is used to overcome the shortcoming of extend kalman filter algorithm which needs the initial value. Also the CLSE is an almost unbiased estimation algorithm. The simulation results show that the CLSE can gradually approach the cramer-rao lower bound and its precision is better than the least square estimation algorithm. Finally, the CLSE is proved to be a gradually, stable and almost unbiased estimation algorithm.