A PSO geomagnetic matching algorithm based on particle constraint

When the initial positioning error of geomagnetic matching navigation is large, due to the vast searching area, particle swarm geomagnetic matching algorithm (PSO) has some problems such as decreasing particle density and low convergence efficiency. In order to speed up the convergence of PSO and increase the geomagnetic matching success ratio, a PSO matching algorithm based on particle constraint is proposed for the geomagnetic navigation. Firstly, the geomagnetic sequence of the matching points is expanded by using the geomagnetic measurement redundant information and combining with the geomagnetic prior map. Secondly, the confidence probability of the real location is calculated and the effective location area is delimited by the constraint of the confidence density function of the search area. Finally, particles are initialized in the effective location area and the particle iteration is carried out to improve the position accuracy of particles. Experiment results show that the particle initialization range under the constraint of geomagnetic redundancy information is 1/4 of the conventional search range with big initial positioning deviation. The proposed algorithm has good matching effect with different measurement noise and better position accuracy compared with other state-of-art algorithms. © 2020, Editorial Department of Journal of Chinese Inertial Technology. All right reserved.