Abstract
To solve the problem of convergence speed and filtering accuracy for conversional extended Kalman filter (EKF) in the research of SINS/GPS integrated navigation system, the application of unscented Kalman filter (UKF) was studied. It approximated a Gaussian distribution by means of deterministically drawed points. UKF replaced the linearization in the EKF by unscented transformation. Compared with the EKF, UKF was more easily implemented and computed efficiently. By means of the computer simulation, two methods were used in the SINS/GPS integrated system, the simulation results showed the UKF overcome the drawbacks of EKF, easily implemented and avoided the Jacobian matrix computation. Compare with EKF, the UKF greatly improve the performance of the states estimate.
| Original language | English |
|---|---|
| Pages (from-to) | 280-283 |
| Number of pages | 4 |
| Journal | Huazhong Keji Daxue Xuebao (Ziran Kexue Ban)/Journal of Huazhong University of Science and Technology (Natural Science Edition) |
| Volume | 37 |
| Issue number | SUPPL. 1 |
| Publication status | Published - Aug 2009 |
Keywords
- Extended Kalman filter
- Global positioning system (GPS)
- Nonlinear filter
- Strapdown inertial navigation
- Unscented Kalman filter