Roll estimation of high rotation speed correction fuze based on extended Kalman filter

Jiawei Wang, Kai Shi, Guotai Xu, Rongzhao Qian, Jie Yan

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

According to the rotating features of correction fuze used on dual-spin stabilized projectile, which is a common designing scheme in 2-D course correction field, a method with Extended Kalman Filter is introduced to solve roll estimation of correction fuze with high rotation speed. Basing on a certain kind of dual spin stabilized mortar projectile, a 7-DOF exterior trajectory model is established to analyse the rotating feature of correcting forward body, and then the analytic formula between roll angle and outputs of gyroscope fixed on fuze is developed and verified. Considering system error of gyroscope, measuring error and BD's positioning error, the real-time roll estimation method is demonstrated using the Extended Kalman Filter method. Simulation results show that the roll estimation absolute error dramatically decreases within 1 s, which is not larger than 6 degrees in whole exterior trajectory. Furthermore, the absolute error in trajectory between 10 s and 40 s is less than 2 degrees. Finally the estimation algorithm is tested in shooting range, the results show that the absolute estimation error is no more than 10 degrees and the mean value of absolute error is 3.9 degrees, which are especially calculated from 15~35 s in whole trajectory. The accuracy of estimation can fully meet require of roll angle measurement in correcting system.

Original languageEnglish
Pages (from-to)938-944
Number of pages7
JournalXibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University
Volume34
Issue number6
StatePublished - 1 Dec 2016

Keywords

  • 2-D course correction
  • Design of experiments
  • Dual-spin stabilized projectile
  • Extended Kalman filter
  • Roll estimation

Fingerprint

Dive into the research topics of 'Roll estimation of high rotation speed correction fuze based on extended Kalman filter'. Together they form a unique fingerprint.

Cite this