Initial alignment is an important question in INS. Obtaining the upmost precision in as short as possible time is the aim in initial phase of INS. Kalman filtering is usually used in initial alignment of INS. Practice improves the fine effectiveness. The essence of using kalman filtering to achieve the initial alignment is that regarding the error angel and the error in inertial instruments as state variables. The error angel and the error in inertial instruments will be estimated using the optimization estimation method. So they are cleared up. The precision and velocity of initial alignment are decided by the precision and velocity of the value of the state variables. But the precision and velocity .of the value of the state variables are decided by the observability of INS. So the observability of INS must be analyzed first of all.In the thesis, it includes: 1. Based on the characteristic of INS, a error mathematics model of "velocity+attitude" matching scheme in transfering alignment is stated. 2. PWCS theory is introduced and used in the error mathematics model of transfering alignment. Using the SVD method and linear system theory , the thesis analyzes the observability of the state variables of INS.The degrees of the observability are given.3. The system is simulated by using kalman filtering. The comparion of the estimate value and the real value is given in figures. And the comparion of the estimate value covariance is also given in figures. 4. The system is analyzed by using the decomposition of constructure. The observational subsystem is obtained and the simulation results are given in figures. 5. Based on the characteristic of INS, a error mathematics model of "velocity+angle velocity" matching... |