Adaptive integrated filter for initial alignment of inertial navigation system

In this paper, a neural network-aided adaptive Kalman/UKF integrated filter was studied for the nonlinear alignment of inertial navigation system. First a more robust learning algorithm for neural network based on the Kalman/UKF integrated filter is derived. Since it gives more accurate estimate of the linkweights, and the convergence performance is improved. This algorithm is then extended further to develop an adaptive Kalman/UKF integrated filter algorithm for state estimation of the nonlinear system. In this algorithm, the neural network is used to approximate the uncertainty of the system noise. And the Kalman/UKF integrated filter is used for both neural network online training and state estimation simultaneously. Simulation results show that the neural network-aided adaptive Kalman/UKF integrated filter algorithm is very effective for nonlinear initial alignment of inertial navigation system. Not only the initial alignment accuracy can surely be obtained, but also the time consumption of each run is reduced considerably. Consequently, an available and practicable algorithm for the nonlinear initial alignment of inertial navigation system is discovered.