Lifting wavelet de-noising for SINS piece-wise fast alignment

A novel piece-wise fast alignment method for strapdown inertial navigation system (SINS) stationary alignment using lifting wavelet is presented. First, the fine horizontal alignment is executed. After the fine horizontal alignment, a lifting wavelet algorithm is utilized to de-noise the output of the gyros, and the azimuth mis-alignment angle is rapidly calculated. Theoretical analysis and practical experiment show that the lifting wavelet algorithm is effective to the de-noising of fiber optic gyro. Compared with the method using Kalman filter in the whole process of the alignment, this method has the same steady-state accuracy, but in the process of the fine azimuth alignment, the mechanical arrangement of SINS is not needed, and the time of fine alignment can be shortened to 30 s or so. This method makes up the defect that the convergence speed of the azimuth misalignment angle is slow, and accelerates the initial alignment process.