20-07-2012, 12:32 PM
Application of Unscented Kalman Filter and Target Tracking
Application of Unscented .ppt (Size: 986.5 KB / Downloads: 75)
Introduction
UAV: An Unmanned Air Vehicle can be remotely piloted or self directing (i.e., autonomous).
Applications:
To simulate enemy aircraft and missile.
Reconnaissance
Combat
Civil and commercial applications: search and rescue, communications relay, research etc
Why UKF?
Other than UKF we have techniques like KF (Kalman filter), EKF ( Extended Kalman filter).
Kalman filter is limited to linear systems. But most of the real world systems are non linear.
Extended Kalman filter attempts to tackle the problem by linearizing the State Vector. But suffers the following limitations-
i. Does not work for considerable non-linearity.
ii. Only Gaussian noise processes are allowed.
iii. Measurement model and dynamic model functions need to be differentiable.
iv. Computation and programming of Jacobian matrices can be quite error prone
Advantages of UKF over EKF
Captures the posterior mean and covariance accurately to the 3rd order (Taylor series expansion) for any nonlinearity
The EKF, in contrast, only achieves first-order accuracy.
Computational complexity of the UKF is the same order as that of the EKF.
Measurement Model in UKF
A simple static camera model is considered.
The terrain is considered to be planar.
Both ground and image coordinate systems have the optical axis passing through their origins.