Underwater geomagnetic navigation based on ICP algorithm

The trend of ocean exploitation puts forward higher requirements for automatic vehicles, while navigation module has been the development constraint for underwater robots. NATURE reported that some sea-turtles recur to geomagnetic "map" for navigation. Although the concrete mechanism has not been made clear yet, this still offers some cue to underwater navigation research. Now bionic researches reckon that matching is one very-possible positioning mode in this subject. Simultaneously remote sensing image processing techniques also supply some inspiration. Its traditional iterative closest point (ICP) algorithm for image registration can be transplanted for location determination. Combined with inertial navigation system (INS) which is as common navigation kernel, this paper emphasizes on how to modify and improve ICP algorithm according to underwater geomagnetic condition. Related aspects such as hunting window concluded from INS and closest points' quick-searching by stepwise extrapolation are also analyzed for the integrity of the proposed method above. The algorithm's applicability is concluded mathematically from the transformation relationship between INS's parameters and related real ones. Then simulation experiments for validation of the proposed algorithm's effectiveness, which is based on the real geomagnetic reference data, are carried out.