Abstract
To improve low nonlinear estimation precision, bad real-time and unstable filtering in trajectory data processing when using unscented Kalman filter (UKF), an adaptive UKF algorithm was proposed. Learning from the idea of strong tracking filer, the gain matrix of adaptive UKF could be adjusted and the trajectory model bias could be compensated by introducing fading factors to modify state prediction covariance matrix online. Moreover, the method of unbiased converted measurement was applied to process radar measurement, which could not only ensure filtering precision, but also simplify the filtering algorithm. The simulation results show that the suggested algorithm has a better performance than the standard UKF, which can be effectively used for real-time trajectory data processing.
Original language | English |
---|---|
Pages (from-to) | 68-71 |
Number of pages | 4 |
Journal | Huazhong Keji Daxue Xuebao (Ziran Kexue Ban)/Journal of Huazhong University of Science and Technology (Natural Science Edition) |
Volume | 42 |
Issue number | 9 |
DOIs | |
State | Published - 1 Sep 2014 |
Keywords
- Adaptive filtering
- Ballisties
- Real-time data processing
- Strong tracking filter
- Unbiased converted measurement
- Unscented Kalman filter