Study On Geomagnetic Matching Algorithm

Geomagnetic Aided Navigation(GAN) is an important military navigation technique by using geomagnetic feature, and it has better application in high tech weapons such as cruise missile, underwater vehicles and so on. Geomagnetic matching is a key technique in GAN. At present, there are some algorithms have been derived for geomagnetic matching which take terrain matching methods as reference. However, as a conservative force field, geomagnetic field is different from terrain field constitutionally. This thesis investigates some new geomagnetic matching algorithms. The main contents and contributions include the following aspects.1. By analyzing the physical characteristics of the geomagnetic field, the thesis introduces the composition of the geomagnetic field briefly, as well as the global models and regional models describing the geomagnetic field. The conception of geomagnetic map and the constructing method are introduced. After that, the thesis builds up geomagnetic matching system and studies geomagnetic matching principle, which sets the basis for the study of geomagnetic algorithms.2. For choosing suitable geomagnetic areas to research geomagnetic matching algorithms, selecting of matching area based on support vector machine is studied. By defining geomagnetic characteristic parameters such as geomagnetic standard variance, fisher information content, code distortion, geomagnetic coarseness and so on, the thesis investigates the method for selecting suitable matching area by using the support vector machine. The design principle of support vector machine is introduced. Then using the measured parameters as input feature vectors of the training set, the thesis pickes suitable kernel function and parameters of kernel function to classify the matching areas and non-matching areas. At last, by using the testing set, the classified effect is tested. This study selected a suitable matching area to analyze geomagnetic matching algorithms.3. Based on describing the feature of geomagnetic signal in frequency and the noise model, the Spatial/Frequency domain Matching(SFM) method is stuied. The process of the algorithm, electing method of the initial searching area, produce of the pending matching locus and the matching strategy in frequency are studied. Using the geodesic geomagnetic data, the SFM method is tested, and the antinoise property, robustness and Real-time property of TFM are analized.4. Affine transformation geomagnetic matching localization based on genetic-searching strategy is researched. This thesis builds up affine transformation model according to the error model of the inertial navigation system, introduces the principle of genetic-searching strategy, and searches the parameters of the affine transformation model based on the strategy. Simulations and experiments are designed to evaluate the algorithm. The results show that the algorithm can solve geomagnetic matching localization satisfactorily, with limiting the position, yaw and velocity errors. The searching efficiency is improved by using the genetic-searching strategy.