A closed-loop procedure for the modeling and tuning of Kalman Filter for FOG INS

This paper describes an iterative closed-loop procedure for the performance optimization of an Inertial Navigation System (INS) based on Fiber Optic Gyro (FOG) technology for airborne applications. The proposed approach focuses on the tuning of the Indirect Kalman Filter (IKF) covariance matrices using inertial sensor errors budgets gathered by the analysis of Allan Variance in a reliable calibration environment. The proposed method identifies a base metrics for predicting the actual filter performance, selecting a suitable combinations of IKF tuning parameters in order to satisfy the system specifications for a particular application. The engineering optimization process spans from the sensor raw data acquisition up to the performance test on the real target HW platform, carrying out the various intermediate steps of algorithm design, simulation runs, code porting and deployment on the embedded INS, lab and on-field testing for performance verification and final comparison of the acquired data output with simulation results to feed the successive tuning iteration. The matching effectiveness between simulated and real data is presented to highlight the beneficial features of this approach.