论文部分内容阅读
在无源跟踪中通常假设目标运动模型为离散白噪声加速度模型,此时状态协方差矩阵奇异,通用的后验克拉美罗限(PCRLB)求解方法失效.另外,观测站位置误差可能对目标无源跟踪性能带来较大影响.对此,首先将状态变量分成非奇异的位置和速度两部分;然后通过构造中间向量得到PCRLB的通用表达式,进而推导了观测站存在位置误差时的PCRLB;最后以多运动站只测角跟踪为例对PCRLB进行了仿真分析,结果表明,所推导的PCRLB计算式可为算法设计提供理论指导.
In passive tracking, it is usually assumed that the target motion model is a discrete white noise acceleration model, in which the state covariance matrix is singular and the common PCRLB solution fails. In addition, Source tracking performance is greatly affected. In this paper, the state variables are firstly divided into two parts: the non-singular position and velocity. Then the general expressions of PCRLB are obtained by constructing the intermediate vectors, and the PCRLB of the observation station is deduced. At last, PCRLB is simulated by multi-sport station only angular tracking. The results show that the PCRLB formula can provide theoretical guidance for algorithm design.