Optimization-based self-alignment method for stationary SINS with geographical latitude uncertainty

Abstract

Abstract To solve the self-alignment problem of strapdown inertial navigation system (SINS) with geographical latitude uncertainty, an optimization-based self-alignment (OSA) method and its improvement for stationary SINS without using the latitude information are proposed. We use only the accelerometer and gyroscope measurements, without the aid of the external latitude information, to determine the Earth rate in the navigation frame. Then we formulate the SINS self-alignment process as a Wahba problem to overcome the disturbances of random noise, and use the estimated Earth rate vector and multiple measurements from the accelerometer and gyroscope to formulate the nonlinear objective function. Moreover, the alignment errors of the OSA method are also presented, based on which we propose a two-position OSA to estimate and compensate the horizontal accelerometer biases to further improve the alignment accuracy. The results of simulation and experiments demonstrate that the proposed OSA method and its improvement perform robust to the noise disturbances and achieve better alignment accuracy than conventional self-alignment methods.