Research Of Information Fusion On Inertial/ Geophysical Integrated Navigation

Integrated navigation is an effective method for improving the holistic performance of the navigation system.The current researches that focus to correct the errors of inertial navigation system with only assistant information was the mainstream in navigation field. With the development of multi-sensor information fusion technology, in the navigation field,it has been the important direction that improving the precision of inertial navigation systems with the assistants of several geophysical information.The research in this thesis focus on the the key techniques of Inertial/Geophysical integrated navigation information fusion.The main contents of this thesis are as follows:(1) The thesis introduces the main characteristics of Inertial/Gravity/Geomagnetic navigation, and according to the well complementarity of inertial and geophysical information,a way integrated patterns of Inertial/Geophysical navigation system is given.(2) According to the weakness of the standard kalman filter can't handle the nonlinear model, the extending kalman filter will consume much more calculation,the scale sampling of unscented kalman filter is easy to produce the nonlocal effect, a self-adaptive and change scale sampling of unscented kalman filter algorithm is forward. Through the simulations result shows that this algorithm can improve the stability of unscented transform.(3) Designs the fuzzy inference system,builds the storeroom of the fuzzy rules and then combine it with the kalman filtering to adjust the covariance matrix of system measurement noise adaptively through the fuzzy control. The simulations show that this algorithm can get a better result in system states optimization estimation,simultaneously it has a superiority of adapting system noise variation and can improve the filter precision.(4) According to the feature of a wavelet can transform separating a noise signal into the signal and noise in real time, the measurement noise covariance is estimated by wavelet transform and used it into fuzzy adaptive kalman filter. The simulations show that this algorithm can effectively avoid kalman filter be prone to invalid in the condition of inaccuracy noise covariance.(5) Base on the federated kalman filter, the filter applied to Inertial/Geophysical integrated navigation information fusion, construct the state equation and measurement equation under Inertial/Geophysical integrated navigation, then the integrated system is simulated,through the result,we find that the integrated system has been improved in precision and stability than any of navigation in the system.Through the analysis in this thesis,it can prove that Inertial/Geophysical integrated navigation system is a system of high reliability and independence.