Vehicle model aided inertial navigation

This paper studies the possibility of improving the accuracy of a low-cost strapdown inertial navigation system mounted on a land vehicle by using a vehicle kinematic model. The inertial navigation system is aided with virtual position measurements, which are predicted by a vehicle kinematic model along with virtual velocity measurements. This data is obtained from the non-holonomic constraints that govern land vehicle motion on a surface. Furthermore, the updated heading information obtained from the INS is used to refine the vehicle kinematic model. Analytic analysis and covariance simulation are carried out for the observability analysis of the aided inertial navigation system. Simulation results prove that the accuracy of the low-grade inertial navigation system aided by a vehicle kinematic model can be considerably improved. The strategies proposed in this paper can be used as a backup navigation loop during outages of GPS for an extended amount of time with the navigation error noticeably bounded. Finally, the redundancy and the integrity of the whole navigation system are also further enhanced.