|Abstract:||In order to enhance the performance of a federated estimation system, a standard closely-coupled Extended Kalman Filter (EKF) has been associated with a magnetic measurement based attitude determination algorithm. The observability of the standard navigation EKF is known to be weakened when the sensor inputs to the filter lack excitation due to low dynamics in trajectory. This weakened observability is due to an inability of the filter to separate attitude errors and sensor biases. However, in much of the previous literature, magnetometers have often been used for both heading stabilization and attitude estimation when other methods were not available. Thus, in this work, a chosen quaternion estimator utilizes measurements from the magnetic field domain, along with a corresponding vector set from an IMU, to generate an attitude solution that aids the EKF. This aiding allows for the filter to observe the attitude errors and separate them from sensor biases that corrupt the system estimates. The theory behind this approach is here described in detail and demonstrated via simulation. Experimental testing with industrial grade sensors was also performed and the results from testing shown. The results from the performed simulations and experiments verify that the aiding provided by magnetic measurements produces improvements in the attitude estimation of the EKF, and this in turn allows for improved estimation of sensor bias. In addition to this, observability analysis shows that magnetometers can be used to increase the observability of the estimation filter.|
Proceedings of the 2016 International Technical Meeting of The Institute of Navigation
January 25 - 28, 2016
Hyatt Regency Monterey
|Pages:||263 - 272|
|Cite this article:||
Morales, Gabriel, Bevly, David, "Observability Analysis of a Modified Closely-Coupled EKF with Magnetic Field Aiding," Proceedings of the 2016 International Technical Meeting of The Institute of Navigation, Monterey, California, January 2016, pp. 263-272.
ION Members/Non-Members: 1 Download Credit