Gene expression programming based matching suitability analysis in geomagnetic aided navigation

In geomagnetic aided navigation (GAN), matching suitability denotes the navigability of candidate matching areas (CMAs) and can be characterized by the suitable-matching features extracted from geomagnetic map. However, the consistency between the single suitable-matching feature and matching probability is not satisfactory. Therefore the suitable-matching features are considered to be synthesized in order to analyze the matching suitability more effectively. In this study, gene expression programming (GEP) is utilized for feature synthesis, and correlation coefficient is treated as the fitness function. Experimental results show that the evolutionary synthetical feature is effective and owns more excellent performance than the single suitable-matching feature. The conclusions of this article can be used for selecting suitable-matching areas and further afford guidance for trajectory planning.