Research Of Characteristics Of Errors In SINS And Methods Of Integrated Initial Alignment

In both military and civil areas Strapdown Inertial Navigation System (SINS) plays such an important role as necessary equipment on vessels, planes, aircraft, and so on. In spite of many advantages such as independence, high precision and robustness, SINS has unavoidable disadvantages. The single use of SINS doesn't meet the needs of practical application, because its traditional independent initial alignment requires a long period and static state. Furthermore, the error in pure inertial navigation system accumulates in the whole period in terms of IMU error, which results in not high enough precision. Thus, it has practical significance to research on characteristics of errors and integrated initial alignment based on SINS.In accordance with the principle of SINS, static and dynamic error equations are deduced and presented in detail. Then we made analysis on the characteristics of errors in static and motion states separately. Three major sources of error stimulate constant and periodic systematic errors, and the motion of vehicle changes the periods.To make clearer the influence of motion and different methods of integrated alignment on the performance and effect of initial alignment, we build several integrated math models with such different observed variables as velocity, position, velocity and position, velocity and attitude. Then we adopt three methods to analyze the observability and made simulations. Due to the differences of coupling degree of states, velocity and attitude mode has the best effect, position mode has the worst effect, velocity and position mode, velocity mode has the modest effect. Another, the fierce motion of vehicle can greatly enhance the observability, even break through the influence of gyro drifts on misalignment angles and increase the precision to a higher level.Considering the imponderable IMU error and low azimuth precision in static state, we bring forward a new method of double-position integrated alignment in the thesis. Simulation results show that such method can estimate gyro drift and azimuth misalignment angle precisely. At last, the availability of KF in integrated initial alignment is tested under static state and motional state through lab experiment. Compared with the traditional control theory, KF reveals its own superiority, thus has great value for practical application.