Research On Gravity Anomaly Filtering Matching Algorithm And Matching Suitability

Aiming at the key technology——matching algorithm as well as the matching suitability of the gravity anomaly map of underwater gravity aided inertial navigation; the research has been mainly made in the dissertation.At first,for huge computational complexity, low matching rate, hysteretic quality and poor real-time of serial relevant extreme matching algorithm,the research has been focused on the gravity anomaly filtering matching algorithm; For the correlation problem between gravity anomaly aided inertial navigation and gravity characteristics, the matching suitability of the gravity anomaly map has been analysed and discussed in this thesis. The main works and innovative ideas are listed as follows:1. A kind of vehicle track simulation has been introduced in the thesis; the method of calculating INS equation based on local navigation coordinate system has been discussed. The INS error equation has been stdudied in detail on the basis of local navigation coordinate system.2. The gravity anomaly Parallel Extend Kalman Filtering (PEKF) matching algorithm has been researched; the rule to choose the best filter based on smoothing and add-weight to the square of residual was simplified. The experiment has been made in area where there are different gravity characteristics and different initial position errors, the result showed that the PEKF algorithm could correct the drift error of INS and improve the accuracy of the navigation system in low initial position error. For the problem of unknown gravity anomaly measurement noise because of the unpredictable gravimetric environment and the disturbance of the measuring instruments et al, the matching algorithm of gravity anomaly filtering based on residual errors to estimate the measurement noise variance adaptively has been put forward; The experiment was done and the result showed this algorithm can not only improve the accuracy of the navigation system in a gross initial position error but also have a better accuracy than PEKF on the same condition.3. On the condition of different Unscented Transform, the gravity anomaly Parallel Unscented Kalman Filtering (PUKF) algorithm has been researched in the dissertation. Aiming at the problem that the false observations might be arose by choosing the same weight coefficient as the UT transformation, the PUKF from the view of probability density to add weight for the sampling data has been put forward according to the relativity of the gravitational field. The experiment has been made with different Unscented Transform and different conditions. The results showed that the problem mentioned above could be overcome based on the correlative probability density Add-weight. It has a better accuracy than PEKF on the same condition.4. The new gravity statistical characteristics have been introduced to analyze the local gravitational field such as the coding and distortion. Gravity anomaly standard deviation, absolute roughness, slope standard deviation, coding and distortion have been calculated in the thesis; the isoclines of some gravity statistical characteristics have been built in the dissertation.5. The perfect matching probability was put forward from the aspect of matching efficiency; the correlation between the local gravity statistical characteristics and perfect matching probability has been researched on the basis of eight-direction. For the problem that the matching suitability of the gravity anomaly map could not been evaluated totally using single gravity anomaly characteristic, a new method on the basis of principal component analysis was come up to analyze the matching suitability of the gravity anomaly map. The comprehensive values of the different area were calculated based on the method above to evaluate the matching suitability of the fitted matching area.6. The platform of gravity anomaly aided inertial navigation and some aided modules have been built in the dissertation.