The Application of Modified Covariance EKF Algorithm to Target-Tracking Modeling

Abstract:

In traditional EKF algorithm, biased prediction is used to calculate the jacobian matrix, which leads to get an inaccurate covariance matrix and influence the estimated performance. This paper applies Modified Covariance EKF to target-tracking modeling to solve the problem. In this algorithm, jacobian matrix is calculated again with state estimation. Through this way, measured values are used to modify the covariance matrix, which makes it more accurate. Consequently, estimated performance is improved.

• Periodical: Applied Mechanics and Materials (Volumes 427-429)
• Pages: 953-956
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.427-429.953
• Online since: September 2013
• Authors: Shu Ai Xu*, Shao Hui Cui, Yuan Zhou, Zhen Bin Tang, Wen Jie Zhu
• Keywords: Extended Kalman Flter (EKF), Modified Covariance, Target Tracking

References

[1] Guoxin Li, Yingjie He and Haitao Hu: submitted to Electronic Design Engineering (2012).

[2] Xuebin Wang, Jianhong Xu and Zhang Zhang: submitted to Computer Applications and Software (2012).

[3] Kinlin Feng: submitted to Taiyuan Teachers College (Natural Science) (2005).

[4] Ju Tan: submitted to Chongqing University of Arts and Science(Natural Science) (2009).

[5] Yongyuan Qin, Hongbin Zhang and Shuhua Wang, in: Kalman Filtering and Navigation Principles, edited by Northwest Industrial University Publishing, Xian (1998).

[6] Zhongkang Sun, Fucheng Guo and Daowang Feng, in: Passive Location and Tracking Technology by Single Observer, edited by National Defense Industry Publishing, Beijing (2008).