基于UKF光电被动目标跟踪及可观性分析

光电经纬仪在跟踪测量过程中经常会由于云层遮挡等原因,导致目标暂时丢失情形,采用数据融合技术是保证系统连续平稳跟踪的一种有效解决办法.由于在数据融合中需要估计目标的状态信息,面临着被动目标跟踪领域普遍存在的非线性估计与可观测性两大难题.采用传统的扩展卡尔曼滤波(EKF)算法会产生较大的估计误差,并易导致滤波发散.介绍无迹卡尔曼滤波器(unscented Kalman:Filter,UKF)来解决非线性估计问题,同时分析了光电经纬仪在实际目标跟踪时的可观测性问题,提出在不可观测条件下保持滤波稳定的方法.Monte-Carlo仿真结果证明此算法有效、可行.

Opto-electro Passive Target Tracking Based on UKF and Its Observability Analysis

During the tracking of the theodolite,targets often lost in the FOV(filed of view)because of the shielding of cloud.An effective method to ensure continuously and smoothly tracking is utilizing the data fusion technology.Due to the status of target need to be estimated in the data fusion,two difficulties which are nonlinear estimation and observability should be solved.The solution of traditional Extended Kalman Filter(EKF)often results in large estimation error and easily diverges.The Unscented Kalman Filter(UKF)is introduced to finish the nonlinear estimation.Meanwhile the observability of opto-electro theodolite is discussed during the actual target tracking,and a kind of method is brought out to maintain filter stability in unobservable situation.The results of Monte-Carlo simulation test that the algorithm is effective and feasible.