In this paper,initial alignment problems of the wander azimuth inertial navigation systemare considered via the pole assignment and observer theory.First,the error equations(1.1)—(1.4)of wander azimuth inertial navigation system aregiven in the geographic coordinates,where the three unit vectors of the axes point to the localnorth,west,and upward directions respectively.In equations(1.1)—(1.4),δV_x and δV_yare the level velocity errors respectively,φ_x and φ_y the errors of the vertical,▽_x and ▽_y theaccelerometer constant errors,ε_x and ε_y the gyro drift rates,g the gravity acceleration,R theradius of the earth,Ω the rate of earth rotation,\phi and \alpha the geodetic latitude and the platf-orm wander angle respectively,and \phi_x and α_c are calculating values of \phi and α.Suppose that the inertial platform keeps in the static state in the alignment process.Thenthe time-invariant linear subsystems ∑_1 and ∑_2 are obtained from equations(1.1)—(1.4),where u_(ij) is control input.For each subsystem ∑_i,the black diagram of the digital exact alignment is shown in Figs.2.1,2.2.The black diagram of the azimuth digital exact alignment is shown in Figs.2.3,2.4.Sometimes alternative level digital exact alignment schemes are designed by the pole assignmentapproach and minimum observer theory,and the azimuth digital exact alignment schemes aredesigned too by the pole assignment approach and full state observer theory.Bilocation alignment technique is used in each azimuth alignment process to estimate theconstant gyro drift rate.