A covariance analysis technique for the federated Kalman filter

A covariance analysis technique for the federated Kalman filter based on suboptimal system models is described. A review of the federated filter is given. Performance techniques for suboptimal filters are reviewed and extended to apply to the federated filter. A simple example is provided to demonstrate the technique. The results are compared with those from a Monte Carlo simulation for validation. The covariance analysis technique is shown to provide equivalent results, and requires substantially less computer time for execution. The covariance analysis solution was obtained in much less time that the Monte Carlo results, allowing for the Kalman filter designer to spend more time evaluating results and less time waiting for simulations to be completed.<
>