基于VSIMM-SRCKF的ADS-B航迹滤波方法研究
2020-04-10杜云,张静怡
杜云,张静怡



摘 要:为了提高ADS-B航迹跟踪精度,并针对交互多模型算法所选模型集而导致的跟踪性能下降的问题,采用基于平方根容积卡尔曼的变结构交互多模型(VSIMM-SRCKF)算法对航迹进行滤波。建立运动目标跟踪的VSIMM 模型集来描述机动目标的系统总模型集合,在滤波过程中,SRCKF递推的更新通过将协方差矩阵开平方得到,使计算复杂度降低,并且使协方差矩阵保持非负定,能够避免滤波中的发散问题。仿真结果表明:提出的基于平方根容积卡尔曼的变结构交互多模型算法(VSIMM-SRCKF)在估计误差均值、估计误差标准差以及平均绝对百分比误差方面均优于IMM-CKF算法和IMM-SRCKF算法,说明VSIMM-SRCKF算法具有更好的跟踪精度,可适应于复杂目标航迹的实时跟踪。
关键词:算法理论;ADS-B;航迹滤波;变结构交互多模型;平方根卡尔曼滤波器
中图分类号:TP273;V355.1 文献标识码:A doi:10.7535/hbgykj.2020yx01004
Research on ADSB track filter method based on VSIMMSRCKF
DU Yun, ZHANG Jingyi
(School of Electrical Engineering, Hebei University of Science and Technology, Shijiazhuang, Hebei 050018, China)
Abstract: In order to improve the tracking accuracy of ADSB, and to address the problem of tracking performance degradation caused by the interactive multimodel algorithm depending on the selected model set. In this paper, the variable structure interactive multiple model (VSIMMSRCKF) algorithm based on the square root volume Kalman is used to filter the flight path. Firstly, a VSIMM model set for moving target tracking is established to describe the total system model set of the maneuvering target. During the filtering process, the recursive update of SRCKF is obtained by squaring the covariance matrix, which reduces the computational complexity and keeps the covariance matrix nonnegative definite, which successfully avoids the divergence problem in filtering. The simulation results show that the variable structure interactive multimodel algorithm (VSIMMSRCKF) based on the square root volume Kalman is better than the IMMCKF algorithm and the IMMSRCKF in terms of the estimated error mean, estimated error standard deviation, and average absolute percentage error. The algorithm shows that the VSIMMSRCKF algorithm has better tracking accuracy and can be adapted to the realtime tracking of complex target tracks.
Keywords:algorithm theory; ADSB; track filter; variable structure interaction multimodel; square root cubature Kalman filter
廣播式自动相关监视(ADSB)是一种基于航空数据链的空管监视技术,与一次和二次雷达监视相比,ADSB 监视建设成本低、范围广、精度高、数据更新快,因此 ADSB 被广泛应用于空管监视和场面监视等领域[13]。但是飞机在向外广播ADSB消息时,一些外部环境因素(如温度、大气、风向等)会给飞机的报文传输和处理造成影响,因此需要通过滤波方法对数据进行预测和修正,使航迹更加准确。
第1期杜云,等:基于VSIMMSRCKF的ADSB航迹滤波方法研究河北工业科技第37卷无迹卡尔曼滤波(unscented kalman filter,UKF)[45]在高维非线性模型收敛性较差;容积卡尔曼滤波(cubature kalman,CKF)[67]在递推过程中,计算量大、数值……
