基于IMM-SCKF-STF的机动目标跟踪算法  

Manoeuvring Target Tracking Algorithm Based on IMM-SCKF-STF

作　　者：李国伟[1] 董跃钧[1] 汤显峰[2] 葛泉波[3] 

机构地区：[1]中原工学院计算机分院,河南郑州450007 [2]浙江大学现代教育技术中心,浙江杭州310027 [3]杭州电子科技大学自动化学院,浙江杭州310018

出　　处：《微电子学与计算机》2012年第10期128-132,共5页Microelectronics & Computer

基　　金：国家自然科学基金(60804064;61172113)

摘　　要：在处理非线性机动目标跟踪问题时,传统的非线性滤波估计算法跟踪误差大且容易引起滤波发散.针对上述问题,研究将强跟踪平方根容积卡尔曼滤波(SCKF-STF)和交互多模型(IMM)算法相结合,提出一种新型的交互多模型强跟踪平方根容积卡尔曼滤波(IMM-SCKF-STF)跟踪算法.该算法在SCKF基础上引入强跟踪渐消因子,使其不仅拥有应对机动目标状态突变的强跟踪能力,同时还具备交互多模型算法的优良机动目标跟踪性能.因此,新算法在机动目标跟踪方面将获得更高的非线性滤波估计精度,且算法的稳定性和应对状态突变的跟踪鲁棒性能获得显著提高.最后,通过两个仿真例子验证了此算法的有效性与优越性.When dealing with tracking problems of nonlinear maneuvering target, the conventional nonlinear filtering methods have larger tracking error and are easy to cause filter＇s divergence. To solve both problems, an interacting multiple model square-root cubature Kalman filter based on Strong tracking filter （IMM-SCKF-STF） is proposed by introducing SCKF-STF into IMM algorithm. In the novel tracking algorithm, it introduces a strong tracking fading factor into SCKF, and makes it possess of not only strong tracking ability to abrupt change of state of maneuver target but also outstanding tracking ability to maneuver target had by IMM method. Accordingly, it has better nonlinear filtering estimate accuracy and algorithm stability for this proposed tracking algorithm, especially the tracking robust performance to deal with abrupt change of state of maneuver target should be improved clearly. Finally, two simulation examples are presented to compare the novel algorithm with the current IMM-UKF on tracking performance and the results show that the novel one is better than IMM-UKF algorithm on estimate accuracy and tracking performance.

关 键 词：机动目标跟踪 非线性滤波 交互式多模型 强跟踪平方根容积卡尔曼滤波 

分 类 号：TN911.72[电子电信—通信与信息系统]