角速度估计自适应的IMM三维目标跟踪算法  

A 3D Turn Rate Estimation Based Adaptive IMM Target Tracking Algorithm

作　　者：田雨芬[1] 蒋宏[1] 丁全心[2] 梁国威[2] 

机构地区：[1]北京航空航天大学控制一体化技术国家级科技重点实验室,北京100191 [2]光电控制技术重点实验室,河南洛阳471009

出　　处：《电光与控制》2012年第10期17-19,29,共4页Electronics Optics & Control

基　　金：航空基金(20095151022)

摘　　要：为了解决三维空间内的机动目标跟踪问题,提出一种基于常速模型和带约束常速率协同转弯模型(CSCT模型)的交互式多模型算法;同时为了解决目标的机动不确定性问题,针对CSCT模型提出一种类似卡尔曼滤波的方法实时精确地估计目标转弯角速度。综合考虑目标角速度的估计值和估计方差,提高目标跟踪的收敛速度,尽可能地减小滤波角速度与机动目标真实角速度之间的误差,最终明显提高了目标跟踪精度。Monte-Carlo 200次仿真结果表明了该算法在性能上有明显提升,对三维机动目标跟踪效果显著,并证实了该方法的实用性。An Interacting Multiple Model （IMM） algorithm based Constant Velocity （CV） model and 3D Constant Speed Coordination Turn （CSCT） model was put forward to solve 3D maneuvering target tracking problem. And a turn-rate estimation method based on similar Kalman Filter was proposed for CSCT model to deal with the uncertainty of target maneuvering. To decrease the constraint of the given information and to obtain a more precise angular velocity as far as possible, the estimated value and the estimate variance of target＇s turn rate were taken into consideration together, the error between the estimated angular velocity and real angular velocity could be decreased during Kalman filtering processing. And the target tracking precision was improved obviously. 200 times of simulation proved the feasibility of the method and showed that the algorithm has better performance for 3D maneuvering target tracking.

关 键 词：机动目标跟踪 3D常速率协同转弯模型 角速度估计 交互式多模型 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计] TN957[电子电信—信号与信息处理]