The Study Of The Information Processing In INS/CNS Integrated Navigation Systems

The error of inertial navigation systems (INS) is cumulated in course of time because of the drift of the measurement unit-gyroscope. Using the celestial navigation system (CNS) and the information processing technology, the paper disscusses the INS/CNS integrated navigation systems. Firstly, the model of gyro's drift time series is constructed .The empirical mode decomposition(EMD) method is introduced to extract the trend item of gyro's drift, but the local mean estimation isn't precision and the end effects exist in the conventional EMD method. The two-tap adaptive time-varying filter and the adaptive border estimation are used to modify the EMD. After the trend extraction, the method of time series analysis is used to make the autoregressive moving average (ARMA) model for stationary time series. To conclude that the gyro's drift model is constituted by the constant and the first order Markov process. The newly development about integrated modes for INS/CNS is introduced, and two modes ,which states are respectively the quaternion error and the gyro's drift, the error quaternion and the gyro's drift, are concluded on basis of the literature and the gyro's drift model. To construct the state equation and the measurement equation. The proper modes are chosen by comparion. Then the paper introduces the kalman filtering and two adaptive filtering methods, and compares and chooses the filtering and calibrate pattern. At last, in the paper the chosen modes are simulated using the kalman filtering and the adaptive filtering, and the filtering divergence is effectively restrained. After comparison, the result gives the proper scheme of integrated systems.