Forward Matching Fusion Algorithm for PDR based on Geomagnetic Fingerprint Map

Aiming at the problem that indoor positioning using pedestrian dead reckoning (PDR) algorithm will lead to error accumulation and the precision of built-in sensors of smartphones is not high enough, indoor positioning research based on the fusion of geomagnetic positioning and PDR positioning algorithm is proposed. A Kriging interpolation based fusion matching algorithm was used to construct geomagnetic fingerprint map to reduce the time spent in data sampling. The fusion positioning is realized by combining the PDR positioning with the hybrid matching algorithm of lattice matching and sequence matching. The improved algorithm solves the problems of global search for geomagnetic sequence matching and PDR positioning error accumulation in traditional methods, improves the positioning accuracy and solves the problems that through the wall or even out of the map caused by the error accumulation. Experimental results show that the maximum positioning error of the proposed algorithm is less than 1m in the complex movement of 120 steps for 60m. The probability of positioning accuracy less than 0.5m is 85%. This can meet the needs of ordinary indoor positioning.