光电跟踪系统几种次优非线性卡尔曼滤波性能比较  

Performance comparison of various suboptimal nonlinear Kalman filter methods for electro-optical tracking system

作　　者：王秋平[1] 陈娟[1] 王显利[2] 王习文[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所 [2]北华大学交通建筑工程学院,吉林吉林132013

出　　处：《电光与控制》2008年第3期10-13,28,共5页Electronics Optics & Control

基　　金：长春光机所三级创新项目

摘　　要：为在光电目标跟踪中选择性能最优的非线性卡尔曼滤波提供参考依据,对光电跟踪系统中应用广泛的扩展卡尔曼滤波、无迹滤波和中心差分滤波3种滤波方法,分别在二阶常速、三阶常加速模型和"当前"统计模型下实现的性能进行仿真实验对比和分析。比较和分析结果显示无迹滤波的性能要明显优于另外两种方法;"当前"统计模型下的滤波精度要高于另外两种模型。In order to provide a reference for selecting optimal nonlinear filter in electro-optical target tracking, the performances of three filtering methods used widely in electro-optical tracking, Extended Kalman Filter（EKF）, Unscented Kalman Filter（UKF） and Central Difference Filter（CDF） are studied. We carried out simulations on performance of the three filtering methods respectively in Constant Velocity（CV） model, Constant Acceleration（CA） model and current statistical model, and also made comparison and analysis. The result illustrated that UKF outperforms the others, and the filter accuracy in current statistical model is better than the others.

关 键 词：光电跟踪 非线性状态估计 扩展卡尔曼滤波 无迹滤波 中心差分滤波 

分 类 号：V556.7[航空宇航科学与技术—人机与环境工程]