An innovative PSO-ICCP matching algorithm for geomagnetic navigation

Geomagnetic aided navigation is an effective approach to eliminate the cumulative error in inertial navigation system (INS). Iterative closest contour point (ICCP) geomagnetic matching algorithm is sensitive to initial error, which will degrade the navigation accuracy greatly. To solve the problem, ICCP and particle swarm optimization (PSO) are improved separately, and an innovative PSO-ICCP geomagnetic matching algorithm is proposed. Firstly, a multi-attribute decision-making problem is introduced to optimize the output of ICCP. Secondly, PSO and ICCP are combined to diminish the sensitive of ICCP to initial error with the global search capability of PSO. Finally, to improve the performance of PSO, a particle initialization strategy is adopted using sliding window and quad-tree. Experimental results verified that PSO-ICCP can effectively reduce INS error. When magnetic measurement noise is 15nT, the position error and heading error are reduced from 1318.83 m, 7.83 degrees to 23.51 m, 2.08 degrees, which is more accurate compared with state-of-art algorithm.