多被动传感器UKF与EKF算法的应用与比较

针对多被动传感器条件下的目标跟踪问题,给出了推广卡尔曼滤波在多被动传感器条件下的具体算法;考虑到多被动传感器目标跟踪需要解决观测非线性的问题,故而将用于非线性系统的基于UT变换的UKF算法应用于所讨论的跟踪问题中,采用检测融合方案,将多个被动传感器的角度观测组合成量测向量,推导了多被动传感器的UKF滤波算法,实现了对目标在三维空间中的全被动跟踪.将两种算法进行了仿真比较,结果表明,采用多被动传感器的UKF算法可以获得比传统的推广卡尔曼滤波算法更为精确的跟踪效果.

Application and comparison of UKF and EKF algorithm in target tracking with multiple passive sensors

The algorithm of EKF suitable for multiple passive sensors is presented.Thinking about the problem of nonlinear measurements in target tracking with multiple passive sensors,the unscented transform based unscented Kalman filter suitable for nonlinear systems is applied to the target tracking with multiple passive sensors.Detection fusion solution is adopted and the bearings measurements from multiple passive sensors are formed into one vector,then the algorithm of UKF suitable for multiple passive sensors is presented.Comparison of two algorithms in passive tracking to a 3D target with multiple passive sensors is illustrated that the tracking precision of UKF-based is higher than that of the traditional EKF-based.