Research On Geomagnetic Matching Aided Inertial Navigation Technology

By measuring the magnetic field data to get the position information and don’tneed to transmit and receive signals, geomagnetic navigation technology gotexcellent confidentiality. The magnetic sensor has the characteristics of smallvolume, light weight and low cost. With the characteristics above, the geomagneticnavigation got important application prospect and great research value. This paperstudied the geomagnetic aided inertial navigation technology based on geomagneticmatching technology, including geomagnetic regional matching suitabilityassessment, geomagnetic matching algorithm and integrated navigation algorithmbased on filtering technique.Firstly, this paper is concerned with the criteria which determine the size ofgeomagnetic matching area and the characteristic parameters of the geomagnetic.By giving a geomagnetic matching probability evaluation method, this papercalculated the matching probability of each region. In order to solve the nonlinearmapping from characteristic parameters to matching probability, this paper gives anonlinear mapping based on BP network and verified it with simulation.Secondly, the geomagnetic matching algorithms are studied. The study of themagnetic contour matching algorithm and iterative closest contour point algorithmsuggests that the former algorithm can only correct the translation error of thereference trajectory and the latter can correct translation and the heading error.Regarding the deficiencies of the above two algorithms when there exist speed errorin reference trajectory, this paper gives a affine transformation algorithm based ongenetic algorithm with the study of genetic algorithm, the error model of referencetrajectory and affine transformation. This algorithm shows great accuracy andefficiency even when there is velocity error in the reference trajectory.Finally, the integrated navigation algorithm which is based on the filteralgorithm is studied. Regarding the nonlinearity of the system, this paper studied theextended Kalman filter algorithm and unscented Kalman filter algorithm. A filterwas established according to the error model of strapdown inertial navigation,matching result and magnetic field value. The effectiveness of the integratednavigation system based on the filtering algorithm is verified by simulation.