Constrained navigation algorithms for strapdown inertial navigation systems with reduced set of sensors

This paper develops a family of algorithms for low-cost strapdown inertial navigation system for land vehicles. Constraints on the motion of land vehicles are defined. They include constraints on vehicle's orientation relative to the Earth surface, and relationship between vehicle's attitude and its velocity direction. Navigation equations are derived that assume validity of these constraints on the vehicle's motion. Compared to standard strapdown inertial navigation, these algorithms reduce navigation errors in the presence of relatively high instrument noise, and at the same reduce number of required inertial sensors. Navigation errors are analyzed and used in an error model for the Kalman filter. Derived algorithms are applied to processing of experimental data.