A sliding-window least-squares estimation method for the biased velocity observation in the inertial-based integrated navigation systems

An effective estimation method is proposed for the constant horizontal velocity biases in the inertial-based integrated navigation systems. First, the analytical effects of the constant horizontal velocity biases on the exact inertial navigation errors are analyzed by the output-correction Kalman filter. Second, the sliding-window least-squares estimation method for the biased horizontal velocity observation is proposed where the INS error characteristic of horizontal velocities is sufficiently utilized. The simulation results illustrate that the estimation errors of constant horizontal velocity biases, which is derived within one Shuler oscillating period, are less than 5%. This estimation accuracy can be adequate in the inertial-based integrated navigation systems.