混合坐标系下基于U变换的水下机动目标被动跟踪

针对仅有角测量的单站无源目标在混合坐标系中进行跟踪所出现的问题,本文提出了应用U变换算法进行改进.该算法采用求取σ点和相应权值的办法,就可以直接得到非线性函数的均值和方差等统计特性,而不需要对非线性函数进行线性化处理.通过仿真,并且与传统的卡尔曼滤波算法进行比较,该算法不仅能够对目标进行很好的跟踪,跟踪误差小,而且运算速度也比扩展卡尔曼滤波算法快.

Bearing-only Tracking for Underwater Target: A Hybrid Coordinate based on Unscented Transform

To solve the problem of the single observer passive target tracking based on hybrid coordinate,we applied the unscented transform algorithm to improve it in the paper.This algorithm works by constructing a set of sigma points and the corresponding weight function.The means and covariance of nonlinear transformation can be achieved directly,without the transformation from nonlinear to linear.Contrasting with EKF,the simulation results show that this algorithm is not only more accurate and less error in tracking but also has a better performance in calculation.