Open Internet by MindsNet

Tuning Extended Kalman Filter for Accurate State Estimation

The author is struggling to properly tune an Extended Kalman filter (EKF) with a MBFS smoother for a project involving 21 states and 6 measurements. Despite using a paper's guidance on cost convergence checks and Q and R scale update formulas, the author faces issues with process residual cost convergence and measurement residual cost validation. The author seeks guidance on addressing these challenges, possibly related to EKF implementation or data issues.

Engineering, Core Engineering, Aerospace Engineering