UnscentedKalmanFilterOD Object

Top  Previous  Next


The UnscentedKalmanFilterOD object provides algorithms for performing Orbit Determination and State Estimation using the Scaled Unscented Transformation and non-linear filtering techniques. Like the traditional Kalman Filter, the UnscentedKalmanFilterOD object allows observations to be processed sequentially, and potentially in real-time. The advantage of an Unscented Kalman Filter is that it propagates the covariance of the state using the full non-linear motion model. A group of representative “sigma points” are selected to approximate the covariance at a given epoch; these sigma points are then propagated through the non-linear force model to the subsequent epoch and the covariance is reconstructed. Furthermore, measurement partial derivatives are not required; the measurement sensitivities are computed based on the predicted measurements for each of the sigma points. This creates a filter that provides more realistic covariance propagation and more robust convergence.  See the Orbit Determination Guide for more information.


Inheritance Hierarchy: ObjectàODEstimationProcessàSequentialFilteràKalmanFilterODàUnscentedKalmanFilterOD


Available In Editions:



Timing Precision Mode

This page describes functionality in nanosecond timing precision mode.


Deprecated Methods


See also

Orbit Determination Guide