One alignment method for SINS in vehicular environment

Self-alignment for SINS is using the measurement of velocity error to adjust the attitude estimate of the system. Since there exists the influence of disturbed movement in the vehicular environment, it is very difficult to get the accurate velocity error information directly. We analyze the influence of several factors on the navigation velocity, and show that the trend of navigation velocity can be used as the measurement of velocity error caused by attitude error. Then based on the estimate of velocity error in the trend method, Kalman filter with additive quaternion error model is adopted for the initial alignment. We compare this method with that directly setting navigation velocity as the velocity error. Through six vehicular experiments, we show that the fine alignment in the trend method can be achieved faster and more precise than comparison method.