Geomagnetic Matching Algorithm Based on the Probabilistic Neural Network

Abstract

In order to realize geomagnetism-aided navigation in the regions with insignificant geomagnetic characteristic, inspired by the pattern recognition idea, the geomagnetic matching problem was transformed into statistical pattern recognition and was analysed from the aspect of pattern recognition. A matching algorithm based on the probabilistic neural network (PNN) was proposed. The selection of training sample for the PNN, the relation between the parameter of the PNN, and the geomagnetic roughness were studied. Then a PNN was constructed for geomagnetic matching. Simulation results show that the proposed method can greatly improve the efficiency and achieve better performance than ordinary correlative matching algorithm. The accuracy of the matching method is less than the unit cell size of the geomagnetic map.