Initial Alignment Algorithm for Platform Inertial Navigation System Based on Fuzzy Kalman Filter

In this paper, we design an initial alignment algorithm for platform inertial navigation system (INS) combining with fuzzy Kalman filter. Firstly, a mathematical model of the initial alignment for platform INS is built. Then the technique, based on the traditional Kalman filter, is employed to the system. Thirdly, the actual covariance of the residual is approximated by the measured value with fading factor, which weights new measurements more heavily than older measurements and is an effective approach to alleviate the influence of noise statistics error. Finally, fuzzy inference system is introduced to update the covariance matrix of Kalman filter measurement noise with exponential convergence. Compared with the traditional methods, the simulation example shows that the proposed strategy can guarantee faster convergence of error.